Charles Sturt University Research Output · Predicting Organisational Citizenship Behaviour ii...

Predicting Organisational Citizenship Behaviour:

More Than Job Satisfaction and Personality

Carolyn Teal

BEd (USA), MProfStudies (Instructional Systems) (UNE),

BSS (Psych) (Honours Class 1) (CSU)

Submitted to Charles Sturt University,

in fulfilment of the requirements for the degree of

Doctorate of Philosophy

Dated: 31 May 2013

Predicting Organisational Citizenship Behaviour ii

Table of Contents

Table of Contents ........................................................................................................... ii

List of Appendices ......................................................................................................... vi

List of Figures ............................................................................................................... vii

List of Tables ................................................................................................................. ix

Certificate of Authorship .............................................................................................. xi

Acknowledgement of Assistance ................................................................................. xii

Ethics Approval – Charles Sturt University ............................................................. xiii

Ethics Approval – Australian Defence Force ........................................................... xiv

Research Approval – Royal Australian Navy ........................................................... xvi

Abstract ....................................................................................................................... xvii

CHAPTER ONE – INTRODUCTION ........................................................................ 1

CHAPTER TWO – AN OVERVIEW OF ORGANISATIONAL

CITIZENSHIP BEHAVIOUR .................................................................................... 14

Defining Organisational Citizenship Behaviour ............................................................ 14

Dimensionality of Organisational Citizenship Behaviour ............................................. 23

OCB Construct Summary .............................................................................................. 40

Predicting Organisational Citizenship Behaviour iii

CHAPTER THREE – PREDICTING ORGANISATIONAL CITIZENSHIP

BEHAVIOUR ............................................................................................................... 43

Job Satisfaction .............................................................................................................. 60

Personality ...................................................................................................................... 70

Five Factor Model of Personality ........................................................................... 74

Five Factor Model of Personality and Job Performance ........................................ 80

Five Factor Model of Personality and Organisational Citizenship Behaviour ...... 82

Summary of Job Satisfaction, Personality, and Organisational Citizenship

Behaviour ....................................................................................................................... 86

Affective Disposition ..................................................................................................... 88

Affective Disposition and Personality ................................................................... 90

Affective Disposition, Job Satisfaction, Personality, and Organisational

Citizenship Behaviour ............................................................................................ 97

Ability .......................................................................................................................... 103

Interpersonal Skills ............................................................................................... 107

Conceptual Model to Predict Organisational Citizenship Behaviour .......................... 113

Building the Model .............................................................................................. 113

CHAPTER FOUR – THREE HUNDRED AND SIXTY DEGREE

FEEDBACK ............................................................................................................... 120

The Development of 360 Degree Feedback in the Working Environment .................. 121

The Technique of 360 Degree Feedback to Develop Personnel .................................. 124

Rating Source Similarity ...................................................................................... 125

OCB Research and 360 Degree Feedback ................................................................... 135

360 Degree Feedback Rater Analysis for the Current Study ....................................... 141

Predicting Organisational Citizenship Behaviour iv

CHAPTER FIVE – METHODOLOGY .................................................................. 144

Participants ................................................................................................................... 146

Materials ....................................................................................................................... 147

Standardised Materials ......................................................................................... 148

General Mental Intelligence ...................................................................................................... 148

Personality. ............................................................................................................................... 149

Positive Affect. ......................................................................................................................... 149

Behavioural Scale Items ....................................................................................... 150

Job Involvement. ....................................................................................................................... 151

Job Satisfaction and Weighted Job Satisfaction. ....................................................................... 153

Organisational Citizenship Behaviour Items ....................................................... 157

Organisational Citizenship Behaviour – Observer Items. ......................................................... 157

Organisational Citizenship Behaviour – Self-Report Items. ..................................................... 161

Interpersonal Skills – Observer and Self-Report Items. ............................................................ 162

Summary of Measures .......................................................................................... 165

Procedure...................................................................................................................... 167

CHAPTER SIX – RESULTS .................................................................................... 169

360 Degree Feedback Rating Source Analysis ............................................................ 170

OCB Rating Source Data Assumption Testing .................................................... 170

Interpersonal Skills Rating Source Data Assumption Testing ............................. 171

Congruence-r Rating Source Analysis ................................................................. 173

Congruence-d Rating Source Analysis ................................................................ 177

Summary of Rating Source Analysis for OCB and Interpersonal Skills ..................... 179

Structural Equation Modeling ...................................................................................... 180

Assumption Testing for SEM ............................................................................... 181

Model Testing ...................................................................................................... 188

Predicting Organisational Citizenship Behaviour v

CHAPTER SEVEN – DISCUSSION ....................................................................... 195

Study Considerations, Limitations, and Possible Future Research Directions ............ 203

Practical Significance of the Research ......................................................................... 209

REFERENCES ........................................................................................................... 214

APPENDICES ............................................................................................................ 230

Predicting Organisational Citizenship Behaviour vi

List of Appendices

Appendix A Organisational Citizenship Behaviour Scale Items from the 1983 Study

Conducted by Bateman and Organ (Organ, Podsakoff, & MacKenzie,

2006, pp.245-246)

Appendix B Major Motivation Theories (Katzell & Thompson, 1990, pp.145-146)

Appendix C Description of the Revised NEO Personality Inventory (NEO PI-R)

Facet Scales (Costa & McCrae, 1992a, pp.16-18)

Appendix D Royal Australian Navy Rank Structure

Appendix E Petty Officer Research Information and Informed Consent Forms

Appendix F Observer Research Information and Informed Consent Forms

Appendix G Behavioural Scale – Self-Report by Petty Officers

Appendix H Colour Coded Behavioural Scale – Self-Report by Petty Officers

Appendix I Behavioural Scale – Observer Report from Supervisors, Peers, and

Subordinates

Appendix J Positive and Negative Affect Scale (PANAS)

Appendix K Organisational Citizenship Behaviour Trial Questionnaire as Assessed

by Category Sponsors

Appendix L Category Sponsor Research Information Sheet

Predicting Organisational Citizenship Behaviour vii

List of Figures

Figure 3.1 Equation representation of the job performance model (Campbell,

McCloy, Oppler, & Sager, 1993, p.43)

Figure 3.2 Integrative model of the job satisfaction-performance relationship.

Replicated from Judge, Thoresen, Bono, and Patton (2001, p. 390).

Figure 3.3 Four level schematic personality structure based on the work of

Eysenck (1998) and Maddi (2001).

Figure 3.4 Personality influences on subjective well-being via affect (Costa &

McCrae, 1980, p.675).

Figure 3.5 Hypothesised correlates of organisational citizenship behaviour versus

in-role task performance, as reproduced in Organ and Ryan (1995, p.

777).

Figure 3.6 Summary of hypothesised pathways for the motivationally-based

variables of personality, positive affect, and job satisfaction.

Figure 3.7 Motowidlo, Borman, and Schmit (1997) theory of individual

differences in task performance and OCB.

Figure 3.8 Summary of casual pathways to predict OCB through ability

(interpersonal skills) using the variables of general mental ability and

personality.

Figure 3.9 Organisational citizenship behaviour prediction framework, as

modified for the work of Organ and Ryan (1995).

Predicting Organisational Citizenship Behaviour viii

Figure 3.10 Initial organisational citizenship behaviour prediction framework for

ability, as measured through interpersonal skills.

Figure 3.11 Initial organisational citizenship behaviour prediction framework for

job involvement, as measured by personality, affective disposition and

job satisfaction.

Figure 3.12 Hypothesised Model: Predicting organisational citizenship behaviour

through interpersonal skills and job involvement pathways.

Interpersonal skills pathway involved: general mental ability,

extraversion, conscientiousness, and agreeableness. Job involvement

pathway involved: extraversion, conscientiousness, positive affect, and

job satisfaction.

Figure 4.1 360 degree feedback: The circle of feedback

Figure 6.1 Final hypothesised model with error terms: Predicting organisational

citizenship behaviour through interpersonal skills and job involvement

pathways.

Figure 6.2 Organisational citizenship behaviour prediction model with

standardised estimates.

Predicting Organisational Citizenship Behaviour ix

List of Tables

Table 2.1 Principal-Factor Analysis with Varimax Rotation for Citizenship

Behaviour

Table 2.2 Comparison of Organisational Citizenship Behaviour Structure

(Podsakoff, MacKenzie, Paine, & Bachrach, 2000)

Table 2.3 Items Included in the Three Dimensional Behavioural Workplace

Questionnaire (Williams & Anderson, 1991)

Table 3.1 Common Job Performance Factors and the Hypothesised Relationships

to OCB and Task Performance (Campbell et al., 1993)

Table 3.2 Performance Determinant Predictors (Campbell, 1990)

Table 3.3 NEO-PI-R Measure of the Five Factor Model (Costa & McCrae,

1992a)

Table 4.1 Summary of Bias Categories for Rating Sources (Fleenor, Smither,

Atwater, Braddy, & Sturm, 2010)

Table 5.1 Job Involvement Items

Table 5.2 Job Satisfaction and Corresponding Job Satisfaction Importance Items

Table 5.3 Final 45 Organisational Citizenship Behaviour Items for Observer

Assessment

Table 5.4 Interpersonal Skills Items for Self-Report and Observer Measures

Table 5.5 Summary of Measures Utilised for Data Collection

Predicting Organisational Citizenship Behaviour x

Table 6.1 Descriptive Statistics and Correlations for the Organisational

Citizenship Behaviour Observers: Self (five items), Supervisors (24

items), Peers (24 items), and Subordinates (24 items) (N=147)

Table 6.2 Descriptive Statistics and Correlations for the Interpersonal Skills

Raters (All Raters - 14 Items, N=147)

Table 6.3 Descriptive Statistics and Correlations for Organisational Citizenship

Behaviour, Interpersonal Skills, General Mental Ability, Personality

Traits, Positive Affect, Job Satisfaction, Weighted Job Satisfaction, and

Job Involvement (N=147)

Table 6.4 Extraversion Facet Correlations with Interpersonal Skills and Positive

Affect

Predicting Organisational Citizenship Behaviour xi

Certificate of Authorship

“I hereby declare that this submission is my own work and that, to the best of my

knowledge and belief, it contains no material previously published or written by another

person nor material which to a substantial extent has been accepted for the award of any

other degree or diploma at Charles Sturt University or any other educational institution,

except where due acknowledgment is made in the thesis. Any contribution made to the

research by colleagues with whom I have worked at Charles Sturt University or

elsewhere during my candidature is fully acknowledged. This thesis reflects the

opinions of the author and not those of the Australian Defence Force or the Royal

Australian Navy.

I agree that this thesis be accessible for the purpose of study and research in accordance

with normal conditions established by the Executive Director, Library Services or

nominee, for the care, loan and reproduction of theses.”

Signed: Date: 31 May 2013

Carolyn Teal

Predicting Organisational Citizenship Behaviour xii

Acknowledgement of Assistance

The phrase “just keep chewing” was continually in my thoughts whilst undertaking this

course of study. I found that the thesis process to be an extremely large undertaking and

there are many people who have assisted me in the process; sometimes creating more

for me to chew on, while others enabling me to chew through their support.

Without the invaluable assistance and support of many wonderful people, this thesis

could never have come to completion. I would like to thank the 202 volunteer Petty

Officers and their supervisors, peers, and subordinates for their time and honesty in

completing the questionnaires. I would also like to thank the Chief and Navy for

enabling me to undertake the study with a Royal Australian Navy population; Her

Majesty’s Australian Ships (HMAS) in the Sydney and Rockingham areas and Royal

Australian Navy establishments across Australia who allowed their personnel to

volunteer during a period of exceptionally high operational activity; and the Category

Sponsors and their staff for completing the Organisational Citizenship Behaviour

relevancy questionnaire.

A special thanks is given to my supervisor, Associate Professor Graham Tyson, who so

willingly provided support and assistance to me during the thesis and encouraged me to

extend my statistical knowledge in the area of structural equation modeling. Graham’s

ability to find time in his extremely busy schedule was always an inspiration to keep on

chewing. I would also like to thank my co-supervisors, Dr Maitland Bowen, Dr Kent

Patrick, and Dr Robert Buckingham for their specialised knowledge and assistance. The

other major support that I received was from the Libraries at CSU and University of

Maryland – the staff was always extremely patient and timely with my requests and I

would like to thank them for all of their assistance and support.

Finally, and with all of my heart, I would like to thank my family – Chris, Emma, and

Alex, and my friends for their endless encouragement, support, and belief in my ability.

I share this work with you.

Predicting Organisational Citizenship Behaviour xiii

Ethics Approval – Charles Sturt University

Predicting Organisational Citizenship Behaviour xiv

Ethics Approval – Australian Defence Force

Predicting Organisational Citizenship Behaviour xv

Predicting Organisational Citizenship Behaviour xvi

Research Approval – Royal Australian Navy

Predicting Organisational Citizenship Behaviour xvii

Abstract

Industrial-Organisational researchers and practitioners continue to advance their

knowledge to provide a competitive edge in today’s dynamic work environment. One

such emerging area over the past forty years that has captured their interest is the

construct of organisational citizenship behaviour (OCB). OCB is a subset of job

performance and involves extra-role behaviour in the workplace. Traditionally, the

investigated antecedents of OCB have been the variables of job satisfaction and

personality. Results of these antecedent studies have been inconsistent, with some

researchers finding small to moderate correlations, while other researchers argue against

the existence of these relationships. One possible reason for the inconsistent results is

that behaviour is complex and usually involves multiple situational and/or individual

predictors. Consequently, researching simple correlational relationships between job

satisfaction-OCB or personality-OCB may not provide a sufficiently thorough picture of

the prediction of OCB.

To address this issue, the current study aimed to advance OCB antecedent research

through the development of a multiple antecedent prediction model. The model

incorporated the traditional OCB antecedents of job satisfaction and personality, while

also including the dispositional variables of positive affect and general mental ability.

These variables were initially integrated into one of two main pathways to predict OCB-

ability (as measured by interpersonal skills) and job involvement.

The final data sample consisted of 147 Petty Officers from the Royal Australian Navy.

Each of these Petty Officers completed self-report measures for personality traits, job

satisfaction, affective disposition, general mental ability, interpersonal skills, job

involvement, and OCB. In addition, each Petty Officer obtained an OCB and

Predicting Organisational Citizenship Behaviour xviii

interpersonal skills assessment from three different sources – a supervisor, a peer, and a

subordinate. The ratings from each of the three sources were initially assessed to

identify the level of similarity between raters sources when using a 360 degree feedback

data collection methodology in OCB research. This analysis was followed by the

assessment of all collected data against a developed OCB prediction model using

structural equation modeling. Goodness-of-fit indices supported the OCB prediction

model, which consisted of the two main pathways – interpersonal skills and job

involvement (2=12.19, 16, p>0.05; GFI=0.98; AGFI=0.96; CFI=1.00; RMSEA=0.00).

The final model demonstrated that interpersonal skills mediated the general mental

ability-OCB, conscientiousness-OCB, and extraversion-OCB relationships.

Extraversion and conscientiousness contributed to both pathways with significant but

small direct pathways to interpersonal skills and positive affect. Job involvement was

observed to mediate the relationships between positive affect-OCB, conscientiousness-

OCB, and job satisfaction-OCB. In addition to a direct pathway between positive affect

and job involvement being observed, the job involvement pathway also demonstrated

partial mediation of the positive affect-job involvement relationship by job satisfaction.

Results indicated that the strongest contribution to the prediction of OCB was

interpersonal skills. This result is important to the area of learning and development

within organisations, as individual difference aspects were shown to make smaller

contributions to predict OCB through the interpersonal skills and job involvement

pathways. It is proposed that it may be possible to develop an individual’s interpersonal

skills through training, thereby producing increased levels of OCB in the workplace.

However, only 13 percent of the variance in interpersonal skills was explained in the

model and further research could investigate other variables that could provide a greater

understanding of the interpersonal skills construct and its contribution to OCB.

Predicting Organisational Citizenship Behaviour 1

CHAPTER ONE – INTRODUCTION

While reading Dennis W Organ’s book titled Organisational Citizenship

Behavior – The Good Soldier Syndrome (1988), I was interested in his personal account

describing the assistance provided by a fellow worker in a summer job many years

prior. The simplified storyline was that the assistance was voluntarily given to Organ,

who greatly appreciated his senior and more experienced colleague’s help, in binding

rolls of paper at the local paper mill, thus preventing a calamity. I began to consider

what personal account I could provide to gain the same level of reader interest. Despite

much consideration, I could not find an example I felt would be appropriate for

inclusion. The examples I considered seemed trivial and not worthy for discussion. I

was disappointed because throughout my professional career, I saw myself as someone

who had been an organisational citizen; voluntarily providing assistance to my

colleagues and the organisation as a whole that went beyond the normal expectations of

my daily duties.

After re-reading Organ’s book (1988) I discovered possible reasons why my

examples may have appeared trivial and mundane. I was attempting to come up with a

single pivotal example, while Organ (1988) described Organisational Citizenship

Behaviour (OCB) as an aggregate of actions rather than a single event. Additionally,

Organ (1988) provided an example of somebody helping him, while I was considering

how I had helped others. To this point, Organ (1988) stated that he had a clear

recollection of the assistance not only because it had averted negative consequences for

him, but he also noted that on the same day he broke his foot, so his memory of that day

was particularly clear. Organ (1988) suggested that his colleague’s memory of the day

maybe less vivid, even forgotten, as the worker may have seen his actions as trivial and


every day. Somewhat comforted, I realised that maybe I was unable to remember my

own individual helping behaviour because it was a pattern of behaviour that was a

normal part of my everyday working life.

I knew that I had voluntarily stayed back late to assist fellow workers with

heavy workloads; I had temporarily employed individuals from the wider organisation

to assist them with issues of confidence and competence despite not being directed to by

my supervisor or actually having a work requirement to do so; I had championed the

decisions of the organisation and supervisors; and I had participated in extra-role

activities and committees that were not required as part of my contractual employment

but did assist the organisation’s functioning and branding. Perhaps after all, I could

consider myself to be a helpful and supportive person in the workforce and that this

behaviour may have increased the effectiveness of my organisation.

The workplace behaviours that I have been discussing are known to Industrial-

Organisational psychologists as belonging to the construct of OCB. In the work

environment it would be fair to conclude that most people have interacted with other

employees that have exhibited similar pro-social working behaviour. These behaviours

include but are not limited to: turning up to work earlier than required; staying back late

when needed; keeping up to date with company policy; attending meetings that were

important to the organisation but were not compulsory to attend; demonstrating courtesy

to other employees and customers; demonstrating a strong work ethic; helping other

members of the organisation when they were sick or have heavy work-loads; and being

supportive of supervisors and colleagues decisions.

OCB has remained an important area for research as it formed part of what could

be argued as the most critical construct in the domain of Industrial Organisational


Psychology; namely job performance. Ultimately job performance is why people are

hired – for employees to perform tasks that contribute to the organisation’s goals. Job

performance and more specific types of job behaviour, including OCB, have gained the

attention of researchers and practitioners alike. Podsakoff, MacKenzie, Paine, and

Bachrach (2000) affirm that OCB was an important construct as these behaviours have

been shown to directly and indirectly affect the bottom line by providing positive effects

to the organisation, to other employees, and customers. As such, OCB research had

provided useful data and theories on which to base many organisational decisions

(Podsakoff et al., 2000).

Prior to further discussion, it is important to identify and define terminology

which has been widely used by Industrial-Organisational psychology literature that may

have a slightly different usage in other domains of psychology. This thesis will

endeavour to utilise the OCB related terminology consistently with the literature. To

more easily comprehend this thesis, the terms of job performance and antecedent

require further clarification. In the literature, job performance is generally referred to as

overall performance in the workplace, and includes subsets such as traditional core

behaviours and OCB. Consequently, when the term ‘job performance’ is utilised, it

refers to overall performance rather than a specific subset unless otherwise mentioned.

The term ‘antecedent’ is generally defined as a predictor that infers causality (Reber,

1995). Subsequently, to identify a variable as an antecedent, the research contains a

methodology to demonstrate causality. However, in OCB research this is generally not

the case. Despite the majority of the literature being cross-sectional and correlational,

particularly studies conducted before 2000, the term antecedent is frequently used.

Noting that OCB is a subset of job performance, the broader construct of job

performance will be briefly considered in conjunction with the relevant research, in an


effort to gain a better understanding of the development of OCB. The broader construct

of job performance can be defined as behaviours or actions that are relevant to the

organisation’s goals and can be measured in terms of the individual’s level of

contribution (Campbell, 1990). All job performance is complex and even the simplest

jobs operate in an environment where the individual requires specific knowledge, skills,

abilities, and attitudes. Campbell proposes that there are three main determinants which

interact to create performance. The first is declarative knowledge, such as facts,

principles, and goals relating to the job. The second involves procedural knowledge and

skill, which is the competency obtained when the individual is able to apply the

declarative knowledge to a given task. This determinant includes cognitive,

psychomotor, self-management, and interpersonal skills. These two components are

generally referred to as the ability of an individual to be able to produce job

performance. Campbell’s third determinant is motivation and involves the choice to

perform, the level of effort, and the persistence of effort. According to Campbell’s

model, the resultant behaviour is based on a requirement for all three determinants to be

present in order to produce effective job performance.

McCloy, Campbell, and Cudeck (1994) tested Campbell’s (1990) model and

demonstrated that all three determinants were indeed necessary to produce job

performance. In other words, for job performance to be effective, the individual must

have a sufficient knowledge of the task, be able to apply that knowledge in the

workplace, while also being adequately motivated to choose to perform the task.

Notably, if one or more determinant is absent, then the task is not completed.

Consequently, the job performance construct is multidimensional, as job behaviour

resulted from a complex interaction of both individual and organisational influences.

When applied to the production of OCB, Campbell’s model requires the relevant ability


(declarative knowledge and procedural knowledge and skills) to be combined with the

motivation to perform.

Performance has been studied in an organisational environment from the early

1900s. Organ (1977; 1988; 1994) stated that the majority of research up until the 1970s,

tended to consider only traditional core ability tasks, rather than more discretionary,

motivationally-based behaviours. Consequently, job performance was observed to relate

to variables of general mental intelligence and specific abilities. Furthermore, utilising

the traditional measures of declarative knowledge and procedural knowledge and skills,

traditional ability was also used to assess the job satisfaction-causes-performance

relationship, which produced disappointing results to say the least. Nevertheless, the

‘common-sense’ assumption that the relationship between job satisfaction and

performance was obvious has created a sustained fascination, ensuring that researchers

have continued to investigate the relationship. Specifically, researchers considered what

specific types of job behaviour were motivationally-based rather than the traditional

perspective of job performance that had been previously studied. These newly identified

motivationally-based behaviours were considered to hold the key to the job satisfaction-

performance relationship. This continued interest led to the development of OCB.

The founding researchers of the OCB construct have included Barnard (1938)

and also Katz and Kahn (1966), while the name and pivotal studies for the structure of

OCB have been credited to Dennis Organ and his colleagues (see Bateman & Organ,

1983; Organ, 1977, 1988; and Smith, Organ, & Near, 1983). In 1977, Organ wrote a

controversial paper contending that despite the scientific evidence to the contrary, the

layman’s belief that job satisfaction caused performance remained a strongly held belief

in the workplace. In the early 1980s, Organ’s students Thomas Bateman and Ann Smith

took up Organ’s 1977 challenge to consider the relationship between job satisfaction


and job performance. Through their work, the construct of OCB was developed when

they started to identify what specific job performance behaviours would be more likely

to relate to job satisfaction, as compared to the traditional forms of job performance that

were usually studied. These early pioneers contended that OCBs were discretionary and

would be related to the mood of the individual. Conversely, core job tasks were the

traditional performance measure and the early job performance researchers observed

they were more strongly related to general mental ability and specific aptitudes.

Consequently, when job performance was researched in the form of OCB, performance

would be related to job satisfaction. In contrast, when job performance was measured in

the form of traditional job behaviours, performance would not be related to job

satisfaction. In addition to job satisfaction, Bateman and Organ (1983) suggested that

environmental aspects and stable individual differences, such as personality traits,

should also be investigated to determine if they had a role in the prediction of OCB.

Organ, Podsakoff, and MacKenzie (2006) argue that personality remains of interest

when researching OCB, as it has been shown to be related to job satisfaction through its

effect on attitudes, and more specifically on affective disposition. They suggest that

personality can predispose the individual to be satisfied; however, situational factors

may counter the predisposition. Nevertheless, it would be likely that personality could

influence the individual’s propensity to be satisfied and produce more OCB.

As time progressed, the results from continued research in conjunction with an

ever changing work environment necessitated the refinement and redefinition of the

OCB construct. Additionally, as a relatively new construct there continues to be a

professional debate over the construct’s title, definition, structure, and even the

methodology to investigate it. The latest definition for OCB is “behavior that is

discretionary, not directly or explicitly recognized by the formal reward system, and in


the aggregate promotes the efficient and effective functioning of the organization”

(Organ et al., 2006, p.3). When using this definition it was noted that the concept of

discretionary was a relative term as it could vary from person to person and situation to

situation. Moreover, Podsakoff et al. (2000) argued that OCBs are increasingly expected

in the workplace by employers, which demonstrated the importance of continued

research of the OCB construct.

While the above definition relates to OCB, there are a multitude of extremely

similar or comparable concepts that have been studied. These include but are not limited

to: contextual performance (e.g. Borman & Motowidlo, 1993, 1997; Borman, White, &

Dorsey, 1995; Motowidlo & Van Scotter, 1994), extra-role behaviour (e.g. Scholl,

Cooper, & McKenna 1987; Van Dyne, Cummings, & McLean Parks, 1995), non-

prescribed behaviour (e.g. Orr, Sackett, & Mercer, 1989), organisational spontaneity

(e.g. George & Brief, 1992; George & Jones, 1997), and prosocial organisational

behaviour (e.g. Brief & Motowidlo, 1986; George, 1990, 1991; George & Bettenhausen,

1990; O’Reilley & Chatman, 1986). These alternative constructs are either part of the

OCB construct or so similar they could be considered the same construct. For the

purpose of this research, these similar constructs were considered equal to OCB or

OCB-related even though they have been labelled differently.

The structure of OCB has been described as multidimensional, which was

formed by an aggregate of a number of specific dimensions (e.g. Organ, 1988; Smith et

al., 1983; Williams & Anderson, 1991). This perspective enabled individual behaviours

to be grouped into numerous dimensions and the aggregate of these dimensions would

be combined to form the overall construct of OCB. Initial work on the OCB structure

was conducted by Smith et al. (1983), who identified the two dimensions of altruism –

behaviour aimed at helping other employees; and generalised compliance – behaviour


aimed to support the organisation. This conceptualisation was further developed by

Organ (1988) who theorised five OCB dimensions of altruism, conscientiousness,

sportsmanship, courtesy, and civic virtue. In the 1990s, OCB was conceptualised as two

dimensions, Organisational Citizenship Behaviours Individual (OCBI) and

Organisational Citizenship Behaviours Organisational (OCBO) (Williams & Anderson,

1991). OCBI were behaviours that benefited individuals in the workplace, such as

helping others or attending meetings on time. Alternatively, OCBO included behaviours

that could benefit the organisation, for example keeping up to date on new

developments and following company rules and procedures.

Contrary to the multidimensional perspective, OCB had been operationalised as

a single latent construct (e.g. Hoffman, Blair, Meriac, & Woehr, 2007; and LePine,

Erez, & Johnson, 2002) in which all OCB behaviours would be representative of a

singular OCB construct. The term latent variable is generally defined as a construct that

cannot be directly measured (Byrne, 2010). As such, OCB were measured through the

observance of the defined behaviours. Researchers such as LePine et al. (2002) and

Hoffman et al. (2007) have argued that because OCB was unidimensional, OCB should

be measured using a representative sample of behaviours. It was generally agreed that

OCB had many related behaviours, although the number of dimensions or if OCB was

actually a unidimensional construct remained contentious. A discussion of the

development of the OCB definition and the debates regarding the structure,

dimensionality, and labelling of OCB are further discussed in Chapter 2.

Chapter 3 discusses both the traditional and more recently studied antecedents of

OCB. As mentioned, attitudes have traditionally been utilised as a predictor for OCB.

One argument for OCB being related to attitudinal factors is grounded in social

exchange theory, which is based on the work conducted by Blau (1964) and Adams


(1965). From an organisational perspective, social exchange theory argues that the

employee is motivated by intrinsic and extrinsic rewards offered by the employer. These

rewards are received when the employee conducts performance that benefits the

organisation. Rather than establishing formal contractual arrangements, the exchange is

founded on a trust relationship between the employer and employee for both parties to

meet their obligations. In comparison, traditional core job behaviours were generally

contracted in a formal manner between the employer and employee. In today’s climate

of employment contracts, it is argued that discretionary behaviour, such as OCB, could

occur more easily based on social exchange theory, as the exchange is not exact and

considered issues of fairness and the opportunity to participate in discretionary

behaviour. While social exchange theory is not specifically an attitudinal theory, it can

be argued that the social exchange is able to influence the related attitude of the

individuals involved. More specifically, if the social exchange is seen as fair, it can

enhance the individual’s attitude, including their job satisfaction. Alternatively if the

social exchange is believed to be unreasonable or negative, a subsequent drop in attitude

would also be expected to occur (Organ et al., 2006).

To investigate the relationship between job performance and attitudinal

behaviours, Organ and colleagues (Bateman & Organ, 1983; Organ, 1977; Smith et al.,

1983) initially considered job satisfaction, which was an attitudinal component of the

employee’s work experience (Locke, 1976). Organ et al. (2006) continued to argue that

it was a reasonable assumption to link positive job attitudes or job satisfaction with

subsequent positive job performance, and more specifically OCB. It is interesting that

research continues to produce mixed results for this relationship with the majority of

meta-analytic studies generally producing only weak to moderate correlations between

job satisfaction and OCB (e.g. Bowling, 2007; Iaffaldano & Muchinsky, 1985; Judge,


Thoresen, Bono, & Patton, 2001; Organ & Ryan, 1995; and Podsakoff et al., 2000).

Researchers continued their efforts to study the job satisfaction-OCB hypothesis.

Further discussion regarding the job satisfaction-OCB relationship is contained at

Chapter 3.

During the same period of time that researchers were investigating the job

satisfaction-OCB relationship, other researchers were investigating the relationships of

personality-overall job performance (e.g. Barrick & Mount, 1991) and personality-OCB

(e.g. Organ & Ryan, 1995). Personality-based research generally considered specific

traits of personality and how they related to job performance. Notably, despite an initial

surge of interest, personality trait theories were considered poor predictors of behaviour

for the first half of the last century. However, over the last 50 years there has been an

increased interest in researching the personality-job performance relationship. The two

main reasons for this increased effort was the development of both the five factor model

of personality and the construct of OCB. The five factor model of personality is

traditionally numbered and labeled as (I) Extraversion, (II) Agreeableness, (III)

Conscientiousness, (IV) Neuroticism, and (V) Openness to Experience (Goldberg,

1990).

With the identification of the five factor model of personality, OCB researchers

began to focus their attention on the role of personality in job performance. Similar to

the findings of the job satisfaction-performance research, the personality-performance

relationship achieved stronger correlations when specific work behaviours, including

OCB, were considered in comparison to overall job performance (e.g. Motowidlo &

Van Scotter, 1994; Organ & Ryan, 1995). Discussion regarding the individual factors of

personality and their reported relationships with both overall job performance and OCB

are contained at Chapter 3.


Over the past 20 years it appears that we have come full circle with OCB

research. While traditional measures of performance considered general mental ability

and specific skills to predict job performance, OCB researchers have begun to consider

what specific skills and mental abilities were required to perform OCB in the

workplace. This research is discussed further in Chapter 3. It is interesting to note that

the research effort studying individual predictor variables of OCB continued in earnest

with the strongest relationships being demonstrated when more than one predictor

variable was included in the prediction model. Notably no single predictor was likely to

be identified as the major predictor due to the complexity of job performance, including

sub types such as OCB. It is important to note that Borman (2004) considered the

predictors of citizenship behaviours were more likely to be dispositional and as such,

the most recent research effort has focused in this area.

In addition to identifying what predictor variables have been studied in OCB

research, it was necessary to consider the development of methodology and how the

current research could contribute to this particular area. Notably one methodological

advancement that has occurred and importantly seems to have the general support of

organisational and management researchers was that self-report measures could produce

results that were confounded with bias from common error variance and social

desirability (Podsakoff & Organ, 1986). In other words, the responses could be

correlated purely because the same person responded to both questionnaires.

Additionally, the respondent may have answered in a pattern that appeared socially

pleasing. Consequently, antecedent studies generally utilised supervisor and/or peer

observer reports when measuring OCB. As the individual could exhibit various

behaviours that were considered as belonging to the OCB construct, it was reasonable to

consider that some behaviour would be more likely to occur in front of specific people


and/or environments. This logic suggested that a range of people observe different types

of OCB. For example a supervisor may be more able to notice behaviour that was

supporting the organisation, while peers and subordinates may observe more

interpersonal helping behaviours. Research conducted by Allen, Barnard, Rush, and

Russell (2000) supported this argument and suggested that the reliability of OCB ratings

was improved when the study considered the rater responses from multiple sources.

When multiple sources from different levels of the organisation are utilised to collect

performance data, the methodology is known as 360 degree feedback. Organ et al.

(2006) further supported the argument that multiple raters should be utilised in OCB

research, specifically indicating that 360 degree feedback may be useful to increase

reliability and interpretability of OCB research. A discussion is contained in Chapter 4,

detailing the development of the 360 degree feedback methodology and the issues

surrounding this approach.

OCB research has continued to thrive, initially focusing on correlational

research. More recently, with the advances in statistical techniques, researchers have

considered causal links between OCB and its possible antecedents. The current study

aimed to extend existing OCB antecedent research through three main areas. Firstly,

while job satisfaction and personality have been useful predictors of OCB, it is argued

that focusing solely of these two variables has limited the current line of investigation. I

contend that the prediction of OCB required more than job satisfaction and personality.

Consequently, this study identified additional variables associated with attitude and

ability to perform OCB. Secondly, this study utilised a 360 degree feedback

methodology to measure the variables of OCB and ability (interpersonal skills).

Analysis was conducted between the raters for these two variables to investigate the


degree of similarity and variance of rater scores. Finally, this study developed and tested

a model of the selected variables to predict OCB utilising structural equation modeling.

To achieve these goals, the current study used the work of previous researchers

to identify the important variables and their relationships with OCB. These variables

were then included in an OCB prediction model. Data were collected for each of the

included variables from a population in the Royal Australian Navy. The rank of Petty

Officer was chosen for the self-report data, while a 360 degree feedback process was

conducted for the measurement of OCB and ability. From an initial sample size of 202

Petty Officers, 147 cases were able to be used to assess an OCB prediction model using

Structural Equation Modeling (SEM) analysis techniques. Chapter 5 details the

methodology of the study and discusses the use of the pre-existing scales and also the

modification or development of new scales utilised in the study. Chapter 6 presents the

results of the 360 degree feedback analysis and the OCB prediction model testing.

Chapter 7 discusses the results observed in Chapter 6 and outlines the limitations of the

current research. In conclusion, consideration is given for possible directions for further

research.


CHAPTER TWO – AN OVERVIEW OF ORGANISATIONAL CITIZENSHIP

BEHAVIOUR

In 1977 Organ wrote a paper titled A Reappraisal and Reinterpretation of the

Satisfaction-Causes-Performance Hypothesis. This paper was attempting to provide a

devil’s advocate position which supported the possibility of the layman’s belief that job

satisfaction caused performance. Organ (1977) considered that researchers, apparently

including himself at times, were all too willing to attempt to disprove what was

considered a common sense relationship. The over exuberance to debunk any layman’s

belief might potentially ignore supporting evidence. Notably, he did not realise at the

time that this article would provide the impetus to a large body of research in the area of

organisational psychology that continues today.

Defining Organisational Citizenship Behaviour

The term citizenship behaviour first appeared in a 1983 study conducted by

Bateman and Organ. Following Organ’s (1977) controversial paper, the authors were

attempting to find evidence to support the layman’s perspective that job satisfaction

caused performance. While traditional measures of performance focused on core work

behaviours that related specifically to each job, these authors considered helping

behaviours as the measure of workplace performance. As there was no previously

established label for the types of helping behaviours that were being studied, Bateman

and Organ (1983) decided on the term citizenship behaviour. Their study utilised a

repeated measure design and obtained a self-report score of the employee’s job

satisfaction from the Job Descriptive Index, developed by Smith, Kendall, and Hulin

(1969). To measure OCB, Bateman and Organ (1983) developed a 30-item


questionnaire which was completed by supervisors. While the OCB scale items were

not cited in their original study article, Organ et al. (2006) provided a copy of the items

which is reproduced at Appendix A. The scale was a composite OCB measure and

measured helping behaviours that included: “compliance, altruism, dependability,

housecleaning, complaints, waste, cooperation, criticism of and arguing with others, and

punctuality” (Bateman & Organ, 1983, p.589). Data were collected twice with a five to

seven week period between data collection points. While their study was unable to

provide evidence of a causal relationship between job satisfaction and OCB, they were

able to demonstrate a moderate positive relationship between these two constructs, as

both the collection points consistently achieved the same correlation of 0.41 (n=77).

More importantly, they provided other scholars with a catch phrase on which to base

future research.

Dennis Organ’s article in 1977, his early research with Thomas Bateman (1983)

and C. Ann Smith and Janet Near (1983), and also a textbook titled Organization

Citizenship Behavior – The Good Soldier Syndrome (1988) are widely recognised as the

initial development work of the OCB construct. However, Organ (1988) acknowledged

although he was credited with the development of OCB, this type of job performance

had been recognised previously in organisations, albeit without directly naming the

pattern of behaviours specifically as OCB. Notably, Organ (1988) cited the work of

Chester Barnard (1938), Daniel Katz and Robert Kahn (1966, 1978), and Fritz

Roethlisberger and William Dickson (1939) as major influences in his development of

the OCB construct.

Organ (1988) contended the OCB construct was first mentioned in literature by

Barnard (1938). Barnard observed the existence of an informal organisation, which

required its personnel to willingly co-operate for the greater good of the company. He


considered that companies were built from a collection of these smaller informal

organisations. Furthermore he argued the structure of the organisation occurred from a

‘bottom-up’ perspective. In other words, the company did not create the smaller

organisations but conversely, the smaller organisations created the company. The

willingness for the smaller, more informal organisation to co-operate was considered by

Barnard to be a result of individual disposition and satisfaction or dissatisfaction in the

workplace. As will be later discussed, this was the initial line of enquiry for Organ

(1977, 1988), Bateman and Organ (1983), and Smith et al. (1983).

Katz and Kahn (1966, 1978) influenced Organ (1988) through their

investigation of supra-role behaviour that enhanced the organisation’s functioning but

did not equate to the traditional view of job performance. In their work, the authors

suggested that effective organisations must attract and retain individuals; ensure these

individuals exhibited role behaviour that was dependable and preferably exceeded the

normal requirement; and additionally, that these individuals exhibited innovative and

spontaneous behaviour which was beyond the normal role requirement. Their argument

implied a sense of citizenship and cooperation for the increased effectiveness of the

organisation, once again providing a sound foundation for Organ and his colleagues.

Organ (1988) was also influenced by the researchers Roethlisberger and

Dickson (1939) through their work on the Hawthorne studies, which are now well

known to most psychologists and organisational management readers. The Hawthorne

studies investigated the relationship between productivity and illumination during a

series of experiments in the Western Electric Company in Chicago between 1924 and

1932. The resultant Hawthorne Effect was generally studied by varying the levels of

workplace lighting in the factory and changing other working conditions, including: the

relocation of workstations, maintaining clean workstations and clearing the floor of


obstacles. However, the resultant gain in productivity was argued to be caused by the

external interest being shown in the workers, thus increasing their motivation to

perform. Notably in their work, Roethlisberger and Dickson (1939) discussed Barnard’s

(1938) distinction between a formal and informal organisation. Specifically, the formal

organisation “includes the systems, policies, rules, and regulations of the plant which

express what the relations of one person to another are supposed to be in order to

achieve effectively the task of technical production” (p. 558). They also noted that an

informal organisation existed which could not be formally recognised by the company

or be accounted for by the formal organisation. This was a group of “sentiments and

values residing in the social organization by means of which individuals or groups of

individuals are informally differentiated, ordered, and integrated” (p. 559).

Utilising the work of Barnard (1938), Roethlisberger and Dickson (1939), and

Katz and Kahn (1966, 1978) amongst others, Organ (1977) began to formalise the

construct of OCB through the consideration of two fundamental questions in the study

of job behaviour and its possible causal relationship with job satisfaction: 1. How did

job satisfaction affect individual behaviour to increase organisational effectiveness? 2.

What discretionary behaviours exhibited by subordinates were desired by managers but

could not be enforced? Organ and colleagues (Bateman & Organ, 1983; Smith et al.,

1983) then began using the term OCB and defined this performance as “individual

behavior that is discretionary, not directly or explicitly recognized by the formal reward

system, and that in the aggregate promotes the effective functioning of the organization”

(Organ, 1988, p.4). In other words, Organ and his colleagues defined OCB as a pattern

of behaviours that enhanced the effectiveness of the organisation but were voluntarily

provided by the employee and could not be forced to be completed or directly rewarded.


The final chapter of Organ’s 1988 book enabled him to discuss the way ahead

for OCB and was aptly titled Unfinished Business. Organ (1988) indicated that OCB

was a relatively new construct and required additional research to refine it. Fortunately,

the amount of subsequent research to investigate and enhance the definition of OCB has

been substantial. Since the initial work conducted by Organ and his colleagues in the

late 1970s and 1980s, there has been an ever increasing interest in OCB research.

Moreover, other researchers have investigated OCB-like job behaviours and despite the

similarity with OCB, they have been alternatively labelled. The construct that remains

most similar to OCB is contextual performance, which was introduced by Borman and

Motowidlo (1993). It is useful to consider its development as research findings for

contextual performance have been aligned with those found for OCB.

Like Organ before them, Borman and Motowidlo (1993) believed job

performance consisted of behaviours that had yet to be identified and formally labelled.

In contrast to Organ (1977) who investigated the role of job satisfaction, Borman and

Motowidlo considered what behaviours contributed to organisational effectiveness.

These later researchers attempted to highlight what job behaviours have been frequently

ignored and thereby not included as assessed job performance criteria for personnel

selection purposes. In addition to the traditional job behaviours that were generally

specific to a job type, named task performance, they identified contextual performance,

which tended not to be role-prescribed. Task performance was defined as behaviours

that “contribute to the organization’s technical core either directly by implementing a

part of its technical process, or indirectly by providing it with needed materials or

services” (Borman & Motowidlo, 1993, p. 73). In other words, the task performance

produced goods and services from base materials, and included selling merchandise,

teaching in a school, and binding paper in a paper factory, just to name a few. In


contrast, contextual performance behaviours “do not support the technical core itself as

much as they support the organizational, social, and psychological environment in

which the technical core must function” (Borman & Motowidlo, 1993, p. 73). Borman

and Motowidlo (1993) suggested that the types of behaviours that could be included as

contextual performance were:

“Volunteering to carry out task activities that are not formally a part of the job

Persisting with extra enthusiasm or effort when necessary to complete own task

activities successfully

Helping and cooperating with others

Following organizational rules and procedures even when personally inconvenient

Endorsing, supporting, and defending organizational objectives” (p. 73)

Utilising the two types of job performance, as identified by Borman and

Motowidlo (1993), Motowidlo and Van Scotter (1994) conducted an investigation to

test the usefulness of the hypothesised distinction between task and contextual

performance. Specifically, the authors attempted to find evidence that task performance

should be distinguished from contextual performance by examining their individual

contribution to overall performance. Their analysis involved investigating relationships

between the two performance constructs, overall performance, training performance,

personality, and experience. Using a sample of 421 United States Air Force mechanics,

the results of hierarchical regression analysis revealed that task performance explained

13% of variance in overall performance above that explained by contextual

performance, while contextual performance explained 11% variance above task

performance. Task performance was positively correlated with experience (r=0.34),

training (r=0.14) and the personality measures of work orientation (r=0.23) and

dependability (r=0.18). Alternatively, contextual performance was found to correlate


with all measures, specifically experience (r=0.16), ability (r=0.15), training (r=0.18),

and the personality measures of work orientation (r=0.36), dominance (r=0.11),

dependability (r=0.31), adjustment (r=0.14), cooperativeness (r=0.22), and internal

control (r=0.26). From their results, the authors concluded that the two types of

performance contributed independently with respect to organisational value, while

different relationship patterns were identified for the variables of experience and

personality. Specifically, task performance correlated more highly with experience,

while contextual performance correlated more highly to personality variables. As such,

Motowidlo and Van Scotter (1994) postulated that task performance and contextual

performance were the two main forms of performance that contributed to overall job

performance and it was useful to distinguish between the two.

Borman and Motowidlo’s (1993) construct of contextual performance was

notably very similar to OCB, as defined by Organ and colleagues (see Bateman &

Organ, 1983; Organ, 1977, 1988; and Smith et al., 1983). One explanation for the

similarity was that Borman and Motowidlo (1993) actually utilised the work of Organ

and colleagues as the basis for their study. However, it remained important that when

considering these two constructs and their similarity, their initial development impetus

was different. Specifically, OCB was focused on job satisfaction-causes-performance

line of enquiry, while contextual performance placed emphasis on personnel selection.

While Organ’s (1988) and Borman and Motowidlo’s (1993) definitions of OCB

and contextual performance were arguably similar, there were also differences. The

most obvious distinction was the concept of discretionary behaviour. The definition of

OCB made implicit reference to the behaviour being volunteered; the employee could

not be punished for failure to carry out the behaviour. Equally, the behaviour was not

directly rewarded if it was undertaken. However, as discussed by Vigoda-Gadot (2007),


organisations and the environment in which they operate are ever evolving and some of

these changes have impacted on the discretionary nature of OCB. While Organ (1988)

originally described OCB as discretionary, organisations have increasingly expected

their employees to undertake OCB in addition to their normal task performance

behaviours (Podsakoff et al., 2000). Consequently, helping and supportive behaviours

have become less discretionary in the workplace, while also being able to be

increasingly rewarded. Borman and Motowidlo (1993) recognised the ever evolving

culture change in organisations and consequently did not require contextual

performance behaviours to be discretionary.

The ongoing cultural change and issues regarding discretionary behaviour and

formal reward structures were also acknowledged by Organ (1997) when he wrote an

article titled Organizational Citizenship Behavior: It’s Construct Clean-up Time. In the

article Organ argued that the OCB definition required refinement due the changes that

the workplace had experienced over time. In 1997 Organ stated that OCB had become

more similar to the construct of contextual performance, as defined by Borman and

Motowidlo (1993), than when it was first defined. As such, Organ (1997) redefined

OCB as performance that provided social and psychological support to the environment

where task performance occurred. This definition of OCB aligned the construct with

that of contextual performance.

In 2006, Organ and colleagues redefined OCB to once again include the term

‘discretionary’ in the description. Organ et al.’s (2006) most current definition defined

OCB as individual behaviour “that is discretionary, not directly or explicitly recognized

by the formal reward system, and in the aggregate promotes the efficient and effective

functioning of the organization” (p.3). This revised definition once again reaffirmed the

requirement for OCB to be discretionary and generally non-rewarded. Although the


later definition allowed that the voluntary behaviour could be implicitly rewarded, for

example keeping your job or being promoted. However, the reward statement was not

required by contextual performance and consequently re-established a conceptual

difference between the two terms, albeit probably in a semantic sense only. Organ

(1997) himself recognised that the concept of discretionary was a relative term as it

could vary from person to person and across different situations. Moreover, as

previously discussed, Podsakoff et al. (2000) argued that OCBs were increasingly

expected by employers. While there continued to be an ongoing debate between the

differences of OCB and contextual performance, there was an acknowledgement in the

wider industrial/organisation community that despite the stated differences, the two

constructs possessed a large degree of overlap (LePine et al., 2002).

In addition to contextual performance, others terms have been researched that

were similar to the OCB construct or were equivalent to a dimension of OCB. In a key

meta-analysis that will be described in more detail in the next section, Podsakoff et al.

(2000) identified over 30 potentially different forms of OCB. From this large pool of

OCB-like behaviours, the major similar constructs included: extra-role behaviour (e.g.

Scholl et al., 1987; Van Dyne et al., 1995), nonprescribed behaviour (e.g. Orr et al.,

1989), organisational spontaneity (e.g. George & Brief, 1992; George & Jones, 1997),

and prosocial organisational behaviour (e.g. Brief & Motowidlo, 1986; George, 1990,

1991; George & Bettenhausen, 1990; O’Reilley & Chatman, 1986). As has been shown,

there have been many terms and concepts that were similar to OCB. For the ease of

comprehension in the current study, all of these constructs will be considered equivalent

to or form a part of OCB and as such, will be termed OCB.

OCB has received ever increasing research attention since its humble beginnings

in 1977 and it continued to be recognised as an important industrial/organisational


construct. Podsakoff et al. (2000) noted that a multitude of management domains (e.g.

human resource management, industrial and labour relations, strategic management,

international business, and leadership) and disciplines (e.g. marketing, hospital and

health administration, community psychology, economics, and military psychology)

have been involved in the impressive growth of OCB research. One of the main focal

points that researchers have considered was the structure of OCB. However, like the

construct of OCB itself, results from this research have produced a range of differing

opinions. Specifically, discussion continues over the dimensionality of OCB, both with

respect to the number of dimensions or if the construct was singular in structure. The

core of this research will now be considered.

Dimensionality of Organisational Citizenship Behaviour

From its very controversial origins, OCB has continued to evolve as a construct,

creating vigorous debate amongst scholars. Like other psychological constructs, the

discussion has not been limited to its definition, as researchers have also debated its

structure. Some scholars argued that OCB was multidimensional (e.g. Organ, 1988,

1997; Organ et al., 2006; Smith et. al., 1983; and Williams & Anderson, 1991) while

others contended it was unidimensional (e.g. Hoffman et al., 2007; and LePine et al.,

2002).

The first known factorisation of OCB was conducted by Smith et al. (1983). To

develop an instrument that measured OCB, the authors utilised a systematic approach

through semi-structured interviews. A group of managers that were not part of the final

study were asked to identify examples of workplace behaviours that were considered

helpful, but not absolutely required for the job. A measure of 16 items was produced

and trialled on a further group of 67 managers. Factor analysis on the results from the


trial indicated the presence of two factors, as compared to the 30-item composite

measure of OCB, as developed by Bateman and Organ (1983).

The OCB measure was then applied to a larger employee population in the

banking industry and their supervisors. Smith et al. (1983) requested the employees to

complete self-report measures for job satisfaction, leader supportiveness, task

interdependence, extraversion and neuroticism, belief in a just world, and demographic

measures. Supervisors were requested to complete the OCB measure. Usable data from

422 respondents representing 58 departments were used in the study. The principal-

factor analysis identified two factors, with a cumulative explained variance of 54.1

percent. The authors’ results of the factor analysis of the 16 items for the OCB measure

are presented at Table 2.1 (Smith et al., 1983). A review was conducted to determine the

extent that each item loaded on to the factors. Items were retained in Factor 1 if the

loading pattern was greater than 0.50 for Factor 1 and less than 0.50 for Factor 2. Factor

2 items were included if the reverse loading pattern occurred. Items not corresponding

to either loading pattern were removed from further analysis. Each factor was scored by

summing the remaining items. Good reliability was observed for both factors (Factor 1:

r=0.88; Factor 2: r=0.85). However, when the relationship was assessed between the

scale scores of the two factors, Smith et al. noted they were strongly related (r=0.45).

Notably, a strong correlation should not occur between two factors that are supposedly

independent. As such, further analysis was conducted to assess the factor structure using

an oblique rotation. The factor structure was replicated, with similar loadings to the

previous analysis. The authors argued that, while the results of the factor analysis

produced similar results, the oblique rotation was supported as a degree of relatedness

between the factors would be theoretically expected. The authors labelled them altruism

and generalised compliance. Altruism related to helping specific individuals in face-to-


face situations, including: helping other employees who had been absent, orienting

others even though it is not compulsory, and helping other colleagues who have heavy

workloads. Generalised compliance consisted of behaviours that were more impersonal

and aimed at helping the work system rather than the individual. Behaviours in the

second factor included being punctual, not taking unnecessary time off work, and not

taking extra breaks.

Table 2.1

Principal-Factor Analysis with Varimax Rotation for Citizenship Behaviour

Item

Factor 1

Altruism

Factor 2

Generalised

Compliance

1 Helps others who have been absent .81 .24

2 Punctuality .23 .61

3 Volunteers for things that are not required .78 .28

4 Takes undeserved breaks ᵃ

.21 .52

5 Orients new people even though it is not required .72 .04

6 Attendance at work is above the norm .21 .59

7 Helps others who have heavy work loads .76 .31

8 Coasts towards the end of the day ᵃ .33 .39

9 Gives advance notice if unable to come to work .22 .52

10 Great deal of time spent with personal phone

conversations ᵃ

.12 .51

11 Does not take unnecessary time off work .07 .62

12 Assists supervisor with his or her work .70 .25

13 Makes innovative suggestions to improve department .76 .08

14 Does not take extra breaks .06 .63

15 Attend functions not required but that help company

image

.39 .34

16 Does not spend time in idle conversation .11 .55

Eigenvalue 5.40 2.17

Percentage variance explained 38.6 15.5

Cumulative percentage of variance explained 38.6 54.1

Note:

Factor loadings of 50 and above are in bold

ᵃ Reversed scoring

Table reproduced from Smith et al. (1983, p.657).


While the initial research of the OCB structure identified two related factors, the

next major development for the dimensionality of OCB was a theoretical amalgamation

of OCB research and related theories. In his influential book, Organ (1988) theorised a

five factor structure of OCB using the findings of Bateman and Organ (1983), Graham

(1991, originally cited in 1986: see footnote1), Podsakoff and Williams (1986), Puffer

(1987) and Smith et al. (1983). Specifically, Organ labelled the OCB dimensions

altruism, conscientiousness, sportsmanship, courtesy, and civic virtue. LePine et al.

(2002) noted the majority of predictor variable-dimension research had utilised Organ’s

(1988) framework and despite being more than 20 years since its conception, it

provided a solid foundation on which to consider the development of the dimensionality

of OCB.

The first two factors in Organ’s five factor model were altruism and

conscientiousness, a modified version of generalised compliance. These factors were a

useful starting point as they were based on the factors identified in the study conducted

by Smith et al. (1983). The first dimension of altruism was well established by the late

1980s. From an OCB perspective, altruism was based largely on the construct

developed by Smith et al. (1983). This form of OCB was noted by Graham (1991) who

labelled a similar behaviour as neighborliness and also Puffer (1987), who used the

term prosocial behavior.

Puffer (1987) considered prosocial behaviour through the investigation of two

non-task behaviours in a retail environment. The first behaviour was a positive

1Graham, J.W. (1986). Organizational citizenship informed by political theory. Unpublished paper

presented at the Annual Meeting of the Academy of Management, Chicago, IL. Attempts to gain a copy of

this paper with the CSU library, the Academy of Management, and the author were unsuccessful. The

author published an article in 1991 that utilised similar concepts as the original work and is cited in this

thesis.


behaviour towards the organisation and labelled prosocial, which was similar to the

dimension of altruism. The author used Brief and Motowidlo’s (1986) definition that it

was behaviour that was: “(a) performed by a member of an organization, (b) directed

toward an individual, group, or organization with whom he or she interacts while

carrying out his or her organizational role, and (c) performed with the intention of

promoting the welfare of the individual, group, or organization toward which it is

directed” (p. 771). In addition, Puffer suggested that prosocial behaviour was

compatible with organisational goals. The second form of behaviour was labelled

noncompliant, which detracted from organisational goals. Examples of this type of

behaviour included breaking the rules or norms, such as: arriving late to work, not

complying with procedures and guidelines, and taking unscheduled work breaks. The

importance of this study was that data were collected from three sources. The data

included objective scores on sales performance and subjective scores for supervisor

ratings on both forms of behaviour and self-report data on independent and

demographics variables. Results suggested that prosocial behavior and noncompliant

behaviour were independent forms of non-task behaviour.

Altruistic job behaviours, as described by Organ (1988), were discretionary and

aimed to help other personnel on an organisationally-based issue or problem. Such

behaviours included helping other personnel who had heavy workloads, lending your

network contact list to another employee, or teaching a novice how to use a computer

software programme even though these behaviours were not part of the formal

description. Organ (1988) noted that while the usual form of altruism was aimed at co-

workers, altruistic behaviour could also include behaviour that helped other people

associated with the organisation, such as customers, clients, suppliers, and venders. The

benefits of altruistic behaviour included promoting good morale between co-workers


and influencing internal and external sentiments towards the organisation. Moreover,

with minimal initial diversion of workplace energy, Organ (1988) suggested this

behaviour could improve capability and thereby increase organisational efficiency.

Notably, Organ et al. (2006) now label this dimension as helping behaviour rather than

altruism, to avoid criticism of the possible motives for the behaviour or a sense of

selflessness.

Conscientiousness was the second dimension identified by Organ (1988) and

was a similar but narrower dimension to generalised compliance, as first described by

Smith et al. (1983). The OCB form of conscientiousness included discretionary

behaviour where the extended effort of the individual went well beyond that which

could normally be expected from employees in the areas of attendance, punctuality,

meeting deadlines, maintaining and enhancing personal cleanliness and neatness,

conforming to rules and regulations, and taking breaks. In contrast to helping

behaviours, conscientiousness was seen as impersonal behaviour where the recipient

was a larger component of the organisation rather than an individual. Organ (1988)

noted that the dimension of conscientiousness was not directed outwardly towards

another person or organisation. Alternatively, the behaviour occurred based on the

individual’s own working code or standard. Conscientious behaviour could benefit the

organisation through the minimisation of absenteeism and subsequent reduction of

administration costs and time that would be required if absenteeism was higher (e.g.

rearranging personnel tasking, hiring additional personnel, training additional

personnel). A conscientious worker was also less likely to waste scarce resources and to

conform to rules and policies such as Occupational Health and Safety regulations,

thereby increasing safety in the workplace and reducing related costs associated with

workplace injury.


The dimensions of conscientiousness and altruism have been shown to correlate

strongly (e.g. Smith et al., 1983). Nevertheless, Organ (1988) continued to consider

these two factors as separate dimensions of OCB. This reasoning acknowledged the

possibility for an individual to be a conscientious worker without necessarily

participating in helping behaviours towards others and vice versa. While Graham (1991)

and Smith et al. (1983) appeared to prefer the term compliance or generalised

compliance due to behaviour such as rule and policy adherence, Organ (1988) argued

that the term compliance suggested obedience rather than being spontaneous and

discretionary. He preferred the term conscientiousness despite acknowledging the

possible confusion this label had with a dimension of the Big Five personality factors

(Organ, 1997).

To build upon Smith et al.’s (1983) two factor structure of OCB, Organ

reviewed additional research with the aim of identifying other important OCB factors.

He subsequently reanalysed the data from Bateman and Organ (1983) and identified a

factor which he named Sportsmanship. While Organ (1988) stated that the reanalysis

had been completed, he did not provide any details regarding the process.

Sportsmanship related to the individual’s discretionary ability to be tolerant and not

complaining during times of stress in the workplace. Individuals who scored highly in

the sportsmanship characteristic generally avoided behaviours such as complaining,

resisting, noting petty grievances, and making ‘mountains out of molehills’.

Sportsmanship behaviour conserved the stamina of management by not providing an

additional issue during the stressful period, thereby allowing management to

constructively attend to the issues or correct the problem. In comparison to helping

behaviours and conscientiousness, sportsmanship was the avoidance of negative


behaviours that would undermine the effectiveness of the organisation (Organ et al.,

2006).

Organ’s (1988) fourth dimension of courtesy was not founded on any empirical

research. Alternatively, it consisted of behaviour that was not included in the first three

dimensions but Organ had still expected to form part of the OCB construct. This form

of discretionary behaviour aimed to prevent conflict or other personal issues in the

workplace. Such behaviours included providing adequate notice of meetings and work

requirements, considering personal circumstances when making decisions that could

affect other personnel, and not discussing personal information about other personnel.

Courtesy contributed to the organisation by reducing the wastage of resources by

providing individuals time to prepare for meetings or a change in work processes.

Additionally, courtesy should be able to reduce frustration and other negative emotions

that could divert an individual’s energy from conducting their work. Organ (1988)

argued that courteous behaviour would overlap with the dimension of helping

behaviour. Nevertheless, these two dimensions can be distinguished through their aims

of prevention versus cure. In others words the courtesy behaviour supports other people

by providing actions to prevent a problem from occurring. In contrast, helping

behaviour provides support to the individual after an issue has occurred.

The final OCB dimension presented by Organ (1988) was labelled civic virtue

which was based on the work of Graham (1991, as originally cited 1986). Civic virtue

related to the individual’s discretionary ability to be involved in, concerned about, and

participating in the organisation. Behaviours in this factor included participating in

committees that were not compulsory, keeping abreast of the development of the

organisation, and promoting the organisation to the community. While there may be

short term losses with respect to productivity, Organ (1988) argued that long term gains


in efficiency would be achieved through ensuring that employees continued to move in

the strategic direction of the organisation. Interestingly civic virtue could sometimes be

seen as non-beneficial to the organisation by management and could result in

disapproval. Consequently, Organ (1988) considered this behaviour to be the most

admirable form of OCB, as it could result in personal cost to the individual.

In addition to the original five dimensions proposed in 1988, Organ proposed

peacemaking and cheerleading could be included as forms of OCB. Peacemaking would

involve providing conflict resolution between two or more parties, while cheerleading

incorporated positively reinforcing the effort of another colleague or team (Organ,

1990a). Notably, these two dimensions have not been utilised widely in mainstream

OCB research (Organ et al., 2006).

Despite the apparent lack of empirical production of independent factors,

Organ’s (1988) five dimensions have been widely utilised as the framework of OCB

research since their inception. It is also important to consider that while some

researchers were attempting to develop an overall construct of OCB (e.g. Organ, 1988;

and Williams & Anderson, 1991); other researchers were investigating specific types of

OCB behaviours (e.g. George & Brief, 1992; Graham, 1991). Podsakoff et al. (2000)

cite more than 30 potentially different forms of OCB have been studied, which had

produced a great deal of existing conceptual overlap between the constructs. To assist in

providing some order to this mass of research, Podsakoff et al. (2000) conducted a

review of research that aimed to explore (a) the conceptual similarities and differences

between the numerous research constructs the have been studied in the literature; (b)

antecedents of OCB behaviour; and (c) the consequences of organisational citizenship

behaviour. Using their knowledge of the literature and pattern identification, rather than

empirical analysis, Podsakoff et al. (2000) summarised the work of significant


researchers in the area of OCB, OCB-like behaviours, or OCB-like constructs and

grouped these behaviours in seven categories. The seven dimensions were labelled as

helping behaviour, sportsmanship, organisational loyalty, organisational compliance,

individual initiative, civic virtue, and self-development. Table 2.2 adapts their work,

summarising the key researchers and the subsequent related OCB research across the

seven areas as described by Podsakoff et al. (2000). As is shown in Table 2.2 and also

in the following discussion, these authors found difficulty in neatly placing the specific

behaviours or factors within these seven dimensions, with overlap existing. This point

reinforced the existence of strong relationships between the key factors and behaviours

theorised and identified in OCB research.

From the perspective of Podsakoff et al. (2000), Helping Behaviours (Table 2.2)

was the first OCB category and has received the greatest research attention and support.

This category has been central to OCB research since its beginning in the early 1980s

(e.g. Borman & Motowidlo, 1993, 1997; George & Brief, 1992; George & Jones, 1997;

Moorman & Blakely, 1995; Organ, 1988, 1990a, 1990b; Smith et al., 1983; Van Scotter

& Motowidlo, 1996; and Williams & Anderson, 1991). Organ’s (1988) altruism (later

renamed helping behaviours by Organ et. al. in 2006) and courtesy factors related to

helping behaviours, along with Organ’s (1990) later behaviours of peacemaking and

cheerleading.


Tab

le 2

.2

Co

mp

ari

son

of

Org

an

isati

on

al

Cit

izen

ship

Beh

avio

ur S

tru

ctu

re (

Pod

sak

off

et

al.

, 2000)

Au

tho

r

Hel

pin

g

Beh

av

iou

r

Sp

ort

sma

nsh

ip

Org

an

isa

tio

nal

Lo

ya

lty

Org

an

isa

tio

nal

Co

mp

lia

nce

Ind

ivid

ua

l

Init

iati

ve

Civ

ic V

irtu

e S

elf

Dev

elo

pm

ent

Sm

ith

, O

rga

n,

&

Nea

r (1

983)

Alt

ruis

m

Gen

eral

ised

Co

mpli

ance

Org

an

(1

988,

199

0a, 1990b

)

Alt

ruis

m

Cou

rtes

y

Pea

cem

akin

g

Chee

rlea

din

g

Sp

ort

sman

ship

C

on

scie

nti

ou

snes

s C

ivic

Vir

tue

Moorm

an

&

Bla

kel

y (

1995

)

Inte

rper

sonal

Hel

pin

g

L

oyal

ty

Bo

ost

eris

m

P

erso

nal

In

du

stry

Ind

ivid

ual

In

itia

tive

Gra

ha

m (

199

1)

Org

anis

atio

nal

Lo

yal

ty

Org

anis

atio

nal

Ob

edie

nce

O

rgan

isat

ion

al

Par

tici

pat

ion

Wil

liam

s &

An

der

son

(1

991)

OC

B-I

O

CB

-O

Geo

rge

& B

rief

(199

2);

Geo

rge

&

Jon

es (

1997)

Hel

pin

g C

o-

work

ers

S

pre

adin

g

Go

od

wil

l

M

akin

g

Co

nst

ruct

ive

Su

gg

esti

ons

Pro

tect

ing

the

Org

anis

atio

n

Dev

elo

pin

g

On

esel

f

Borm

an

&

Mo

tow

idlo

(19

93,

199

7)

Hel

pin

g a

nd

Cooper

atin

g w

ith

Oth

ers

Hel

pin

g a

nd

Co

oper

atin

g w

ith

Oth

ers

En

do

rsin

g,

Sup

po

rtin

g,

and

Def

end

ing

Org

anis

atio

nal

Ob

ject

ives

Fo

llo

win

g

Org

anis

atio

nal

Ru

les

and

Pro

ced

ure

s

Per

sist

ing

wit

h

En

thu

sias

m a

nd

Ex

tra

Eff

ort

Vo

lun

teer

ing

to

Car

ry O

ut

Tas

k

Act

ivit

ies

Va

n S

cott

er &

Mo

tow

idlo

(19

96)

Inte

rper

sonal

Fac

ilit

atio

n

Job D

edic

atio

n

Job D

edic

atio

n


Organ’s (1988) construct of sportsmanship related to Podsakoff et al. (2000)

second grouping, also named Sportsmanship. This grouping included the sportsmanship

behaviour as cited by Organ (1988) in addition to behaviour that maintained a positive

attitude; was less likely to be offended or take things personally when others did not

follow their suggestions or when their ideas were rejected; and were willing to sacrifice

personal gain for the overall good of the organisation. Notably, the helping and co-

operating behaviours as developed by Borman and Motowidlo (1993, 1997) were

included in Podsakoff et al. (2000) first and second behaviour groupings of helping

behaviours and sportsmanship.

Organisational Loyalty included behaviours that promoted and protected the

organisation. Podsakoff et al. (2000) noted this factor had received inconsistent results

to confirm it as a distinct factor of OCB (e.g. Moorman & Blakely, 1995; Moorman,

Blakely, & Niehoff, 1998) and subsequently required additional research attention.

Alternatively, Organisational Compliance maintained a prominent focus in OCB

research. This grouping related to the Smith et al. (1983) construct of generalised

compliance and included behaviours that related to the individual accepting and

following the organisation’s rules and procedures.

While Organ (1988) based conscientiousness on the generalised compliance

factor, Podsakoff et al. (2000) included this form of behaviour in the Individual

Initiative category. Individual Initiative has received a large amount of research

attention as a major factor of OCB (e.g. Borman & Motowidlo, 1993, 1997; George &

Brief, 1992; George & Jones, 1997; Graham, 1991; Moorman & Blakely, 1995; and

Van Scotter & Motowidlo, 1996) and included job dedication behaviours that go

beyond what is generally expected. Additionally, Podsakoff et al. (2000) labelled Civic

Virtue a dimension of OCB, which related directly to Organ’s (1988) dimension of the


same name. This later dimension included behaviours aimed at protecting the

organisation and also the level of participation in the organisation (see George & Brief,

1992; and Graham, 1991, respectively).

The final grouping of Self Development (Podsakoff et al., 2000) was also not

included in Organ’s (1988) work. George and Brief (1992) developed this dimension

from the work of Katz (1964) and included discretionary behaviours that aimed to

improve the individual’s knowledge, skills, and abilities. Behaviours included attending

non-compulsory training, keeping abreast of the latest developments in their field of

expertise, and expanding or developing new skill sets.

While the aim of Podsakoff et al. (2000) was to provide clarity of researched

OCB constructs, it appeared that consistent with other researchers, they had managed to

create new names for OCB factors, some of which have yet to provide sufficient

empirical evidence to support them. For example, while Podsakoff et al. (2000) used the

title of organisational compliance, they did not include conscientiousness as part of this

category. Alternatively conscientiousness was categorised under individual initiative,

apparently ignoring Organ’s (1988) definition surrounding the compliance with rules. It

could be argued that organisational loyalty actually consisted of behaviours that formed

part of both the sportsmanship and organisational compliance dimensions. Additionally,

self-development could contain behaviours that were included in a conscientiousness

factor as the individual was increasing their current competency and capabilities. As

these behaviours could be theoretically included in Organ’s (1988) dimensions, it is not

surprising that factorial investigations have produced inconsistent results.

While Organ’s 1988 five factor taxonomy was an important and founding OCB

structure, an alternative approach was proposed by Williams and Anderson (1991) who


conceptualised OCB as two dimensions. Notably, their work has provided the basis for

a considerable amount of research and is also shown in Table 2.2. Their study aimed to

test a three dimensional performance model and was based on the work of Bateman and

Organ (1983), Graham (1991, as originally cited in 1986), Organ (1988), and Smith et

al. (1983). The dimensions included behaviours that were directed at other people

(OCBI), behaviours that were directed at the organisation (OCBO), and in-role

behaviours (IRB). The scale for their work is contained at Table 2.3.

Table 2.3

Items Included in the Three Dimensional Behavioural Workplace Questionnaire

(Williams & Anderson, 1991)

OCBI OCBO IRB

Helps others who have been

absent.

Attendance at work is above the

norm.

Adequately completes assigned

duties.

Helps others who have heavy

workloads.

Gives advance notice when

unable to come to work.

Fulfills responsibilities specified

in job description.

Assists supervisor with his/her

work (when not asked).

Takes undeserved work breaks.

(R)

Performs tasks that are expected

of him/her.

Takes time to listen to co-

workers’ problems and worries.

Great deal of time spent with

personal phone conversations.

(R)

Meets formal performance

requirements of the job.

Goes out of his way to help new

employees.

Complains about insignificant

things at work. (R)

Engages in activities that will

directly affect his/her

performance evaluation.

Takes a personal interest in

other employees.

Conserves and protects

organizational property.

Neglects aspects of the job

he/she is obligated to perform.

(R)

Passes along information to co-

workers.

Adheres to informal rules

devised to maintain order.

Fails to perform essential duties.

(R)

Williams and Anderson (1991) collected performance data from 127 fulltime

employees and their supervisors and peers. The authors suggested their results achieved

a three factor solution for performance with IRB, OCBI and OCBO. Similar to the

results gained by Smith et al. (1983), the authors found that the two OCB type factors


were strongly related (r=0.56). Additionally, there were strong intercorrelations

between the IRB and both OCBI (r=0.52) and OCBO (r=0.55). OCBI behaviours

included Organ’s 1988 dimensions of altruism or helping behaviour and courtesy in

addition to peacemaking and cheerleading; while OCBO encompassed the three

remaining 1988 dimensions of sportsmanship, conscientiousness, and civic virtue.

The previous discussion has highlighted the difficulties researchers have had

attempting to find consistent results when identifying dimensions. While it could be

argued that there has been an over-reliance of intuition and professional judgement

rather than the use of empirical studies, a further issue related to the assumption that

OCB was actually multidimensional. In other words, OCB was formed from a number

of different groups of behaviours. As such, overall OCB would be the summation of all

of the OCB dimensions. Notably, Organ has maintained a dimensional perspective of

OCB throughout his career (Organ et al., 2006). Perhaps the majority of OCB research

has utilised this approach because of Organ’s influence in this area. He was the founder

of the OCB construct and is correctly viewed as the subject matter expert in the field.

However, given the results have been inconsistent in providing a distinct set of

dimensions, other structures should be considered. One such possibility was that OCB

was a unidimensional variable, which suggested that all OCB individual behaviours

were imperfect indicators of a single construct.

The meta-analytic studies conducted by LePine et al. (2002) and Hoffman et al.

(2007) are considered pivotal investigations that have researched the OCB construct

from a unidimensional perspective. In the earlier study, LePine et al. (2002) recognised

the conceptual overlap that existed between the previously established/theorised

dimensions, as shown at Table 2.2. With a willingness to enhance the empirical

knowledge of this issue, their research focused on the structure of OCB. Specifically


they wanted to investigate if the structure of OCB was multidimensional, such as the

five factor framework proposed by Organ (e.g. Organ, 1988; 1997; Organ et al., 2006);

bi-dimensional where the elements are sub-grouped into two main dimensions, such as

that proposed by Williams and Anderson (1991); or perhaps the OCB construct was

unidimensional, where a range of behaviours could be measured as imperfect predictors

of a general tendency to perform OCB in the workplace. Studies were identified for

inclusion for a meta-analytic approach from a literature search in PsycINFO database

between the years of 1983 to 1999 and directly contacting known researchers of OCB.

LePine et al. (2002) selected Organ’s five dimensions from a variety of different

approaches that had been utilised to investigate the dimensionality of OCB. Their

reasoning was that Organ’s dimensions were used by the majority of OCB researchers

and produced the most consistent framework over an extended period of time. The

authors also stated that previous research had identified Organ’s five dimensions

utilising factor analysis. However, LePine et al. argued that additional analysis was

required to determine if the five hypothesised OCB dimensions would be related to each

other, and secondly, if they demonstrated similar or different patterns of relationships

with the five common predictors of satisfaction, commitment, fairness, leader support,

and conscientiousness. Results from their meta-analysis of 37 studies indicated that

while the five dimensions had been identified using factor analysis, further analysis

between the dimensions and also with the five main predictors did not provide evidence

to suggest that OCB was a multidimensional construct. Specifically, the factors were

shown to be highly related, similar to the results of Williams and Anderson (1991) and

Smith et al. (1983). Moreover, LePine et al. demonstrated that there were no meaningful

differences between the five common predictors and the five OCB dimensions. These

results suggested that satisfaction, commitment, fairness, leader support, and


conscientiousness were able to predict OCB, independent of how OCB was defined. As

such, the authors suggested that the OCB dimensions should be considered as imperfect

indicators of OCB. Given their findings, LePine et al. recommended that other OCB

researchers would be better served to view OCB as a general tendency to be helpful and

cooperative in a work setting rather than to focus on specific dimensions.

Hoffman et al. (2007) also investigated the dimensionality of OCB through

meta-analytic methodology, specifically considering Organ’s five dimensions. The

authors conducted analysis on 112 studies to examine the dimensionality of OCB, the

distinction between OCB and task performance, and the relationship between OCB, task

performance, and attitudinal variables. Using confirmatory factor analysis, the study

modelled the two-dimensional OCB structure, which corresponded to the OCB

conceptualisation by Williams and Anderson (1991) against Organ’s five dimensions.

Additionally, the model included the following six latent variables: task performance,

job satisfaction, organisational commitment, distributive justice, interactional justice,

and procedural justice. While the model provided a good fit to the data, the two OCB

factors were considered to be measuring the same construct as they were highly

correlated (r=0.98). Subsequently an overall measure of OCB was modelled with the

other six variables, corresponding to the unidimensional conceptualisation of LePine et

al. (2002). The second model supported the results gained by LePine et al. (2002),

providing further evidence that OCB should be considered as a unidimensional variable.

Given the results of LePine et al. (2002) and Hoffman et al. (2007), one could

expect the conceptualisation of OCB to be unidimensional. As with other aspects of

OCB, the debate continued. Marinova, Moon and Van Dyne (2010) conversely argued

that the OCB construct was increasingly complex and suggested that the measurement

of OCB remained overly simplistic. In an effort to conceptualise the structure of OCB,


the authors utilised Williams and Anderson’s (1991) conceptualisation of OCB using

two types of behaviour that are aimed to benefit the individual versus the organisation.

In addition, their study considered two other types of OCB behaviours – promotive

versus protective. Promotive OCBs were categorised by an ability to be adaptive and

also could involve behaviours that involved proactive helping of either the individual or

the organisation. Alternatively protective OCB included behaviours that aimed to

protect either the individual or organisation by maintaining stability and predictability.

Using these different types of OCB, Marinova et al. considered the four previously

discussed types of OCB of helping, sportsmanship, compliance, and taking charge.

Helping behaviour was considered to be a combination of interpersonal and promotive

OCB, while sportsmanship was interpersonal and protective OCB. Compliance was the

combination of organisational and protective, while taking charge was viewed as a

grouping of organisational and promotive OCB. Results using factor analysis and

structural equation modeling provided evidence for a four factor solution. Furthermore,

structural equation modeling demonstrated that the model provided a better fit to the

data for a four factor model solution, as compared to either a one factor (e.g. Hoffman et

al., 2007; LePine et al., 2002) or a two factor model (individual versus organisation;

promotive versus protective). This study demonstrated that while there were many more

dichotomies that could be utilised than the two presented in their study, the construct of

OCB was complex and its dimensionality remained of interest to researchers.

OCB Construct Summary

Chapter 2 has provided an overview of the development of the OCB construct

with respect to its definition and structure. As has been demonstrated, both areas remain

contentious and have continued to be debated since the first controversial paper by


Organ (1977). Organ, at that time, was unaware of the significance his paper was to

have on the research community. However, following his work and that of his colleges

in the 1980s, Thomas Bateman, C. Ann Smith, and Janet Near, Organ started a body of

OCB research that continues with enthusiasm today.

The latest definition from Organ was contained in his book (2006) that he co-

authored with Podsakoff and MacKenzie titled Organizational Citizenship Behavior –

Its Nature, Antecedents, and Consequences. He has returned to a more traditional

position of OCB which is “Individual behavior that is discretionary, not directly or

explicitly recognized by the formal reward system, and in the aggregate promotes the

efficient and effective functioning of the organization” (p.3).

The amount of OCB research has been ever increasing since the 1980s resulting

in construct confusion. Researchers have studied similar constructs, OCB as a whole,

and separate dimensions. While Podsakoff et al. (2000) cited more than 30 such

constructs or dimensions, just two years later in another key meta-analytic study,

LePine et al. (2002) identified more than 40 constructs. It was important to reiterate that

numerous constructs existed and for the purposes of this study, all of these constructs

were considered equivalent and would be termed as OCB.

Differing viewpoints regarding conceptual overlap and the dimensionality of

OCB continue today. The structure of OCB is studied as both multidimensional and

unidimensional constructs. The issue of structure continued to interest researchers to

conduct further research and apply professional judgments to add to the collected

knowledge. It should be noted that while dimensions have been identified, they are

highly related to each other and had similar antecedents (LePine et al., 2002). Moreover,

while Organ had produced the factors of altruism and conscientiousness statistically,


Organ’s five dimension structure was generally theorised rather than statistically

derived. Despite the apparent lack of scientific determination, OCB researchers

appeared to readily support the dimensions theorised by Organ and his colleagues

(1988, 2006).

It was not surprising that a relatively new construct, such as OCB required

continued research effort and refinement. As a construct, it operated in a constantly

changing workplace that continued to seek a competitive edge. It was important that a

scientific approach continued to be utilised to enable the investigation of OCB construct

definitions and structural viewpoints. While this dimensionality research occurred

perhaps we should take the advice of LePine et al. (2002). That advice suggested that

we may be better served to consider the OCB construct as a whole, rather than

measuring individual dimensions. They also suggested that it was meaningless to

conduct research on individual dimensions and researchers should consider a complete

measure of OCB. Consequently, for the purpose of this current research, I will consider

OCB to be a unidimensional construct that could be measured using a range of

behaviours, including the five types of behaviours, as described by Organ et al. (2006).

In the next chapter, I will discuss the main antecedents of OCB that have been

investigated. The initial antecedent work and impetus for the development of OCB was

the job satisfaction-OCB relationship (Organ, 1977). Over time, a second antecedent

had emerged, namely personality. This new line of research was one of the most

interesting areas because the five factor model of personality appeared to have obvious

relationships with OCB. However, as the title of the thesis suggested, the prediction of

OCB was more complex than could be explained by job satisfaction and personality

alone. Subsequently other emerging variables will also be discussed in the following

chapter with the aim of providing a more complete picture in the prediction of OCB.


CHAPTER THREE – PREDICTING ORGANISATIONAL CITIZENSHIP

BEHAVIOUR

Job performance has remained one of the most important constructs for

Industrial-Organisational psychologists and practitioners alike (Borman, 2004).

Research investigating job performance was based initially on a singular performance

dimension. Specifically, during the first half of the last century, performance research

focused on traditional core job behaviours (Fay & Sonnentag, 2010). This type of

performance was linked to workplace productivity and was produced to achieve the

organisation’s main objectives. Behaviours such as: a pilot flying a plane; a doctor

prescribing medicine for a patient’s illness; an electrical technician wiring a house; or a

plumber checking for blockages in a drain are all examples of core behaviours that have

been traditionally used to measure job performance. As previously discussed in Chapter

2, Borman and Motowidlo (1993) coined the term task performance, which was

synonymous with these traditional core job behaviours.

Initial performance prediction research suggested that the best predictors of task

performance were general mental ability or g; and specific abilities or s (e.g. Spearman,

1904). There has been a large body of work that investigated what specific abilities

would increase the prediction of task performance. As a result of this continued effort,

many aptitudes have been identified. Doverspike, Cober, and Arthur (2004) stated that

common aptitudes have included general intelligence, verbal ability, numerical ability,

spatial ability, mechanical comprehension, memory, perceptual speed, psychomotor

ability, adaptation ability, and social intelligence. However, more recently, the

complexity of job performance has captured researchers’ attention. Over time, the

performance construct has evolved to become broader, which has enabled the


consideration of additional types of job behaviour. As such, researchers have also

investigated what antecedents would predict these new forms of performance, including

OCB.

The current study researched various theoretical performance models to be able

to identify OCB antecedents. It is not surprising given the importance of the construct,

that many job performance models have been developed. One of the most commonly

cited performance models was developed by Campbell (1990). His model has been used

to investigate both the core job behaviours that were initially studied and also other

more recently proposed forms, such as OCB. In his model, Campbell contends that job

performance consists of specific behaviours. These behaviours are performed by the

employee and are relevant to achieving organisational goals. Furthermore, Campbell

emphasises that job performance is not the consequence or result of the behaviour, but

rather the action itself.

Campbell’s (1990) performance model posits that all jobs, independent of the

level of task complexity, consist of a variety of behaviours termed performance

components. A performance component represents a large task within each job and can

be specific to the job or generic across different jobs. Campbell, McCloy, Oppler, and

Sager (1993) extended Campbell’s 1990 model by describing an eight-factor structure

that represented the major performance components of the majority of jobs. The eight

factors are: job specific task proficiency; non-job-specific task proficiency; written and

oral communication task proficiency; demonstrating effort; maintaining personal

discipline; facilitating peer and team performance; supervision/leadership; and

management/ administration. A description of each of the eight performance

components is detailed in Table 3.1. Campbell et al. note that while not all jobs


contained all eight aspects, they were useful to be considered as latent variables for all

types of job performance.

Table 3.1.

Common Job Performance Factors and the Hypothesised Relationships to OCB

and Task Performance (Campbell et al., 1993)

Common Job

Performance Factors

Description Relationship to OCB and

Task Performance

1 Job-specific task

proficiency

Core substantive or technical tasks that

are job-specific. These behaviours

distinguish one job from another.

Task performance

2 Non-job-specific

task proficiency

Performance behaviour that is not

specific to a particular job.

OCB and non-specific task

performance.

3 Written and oral

communication

task proficiency

Formal oral and written presentations to

audiences that involve writing and

speaking.

Aspects of interpersonal

facilitation can be applied to

both task performance and

OCB

4 Demonstrating

effort

The degree the individual is committed

to job tasks, works at a high level of

intensity, and keeps working despite the

working conditions.

Motivational aspects can be

applied to both task

performance and OCB.

5 Maintaining

personal discipline

The degree to which negative behavior

is avoided.

Motivational aspects can be

applied to OCB

6 Facilitating peer

and team

performance

The degree to which the employee

supports other employees and the

workgroup through problem solving,

role modelling, team facilitation,

completing work goals, and informally

training others.




OCB

7 Supervision and

leadership

Interpersonal behaviours that influence

the performance of the subordinate

through goal setting, training, modelling

competent performance, punishment,

and reward.




OCB

8 Management and

administration

Behaviours that are directed at

workgroup goal setting, monitoring

progress, problem solving, financial

management, resource allocation, and

representation.




OCB

The eight performance components, as presented by Campbell et al. (1993),

provide a useful framework to analyse all forms of performance. Table 3.1 suggests

how OCB and task performance could be expected to relate to the performance

components. I argue that OCB does not contain job-specific task proficiency, as this


behaviour would equate to task performance. In contrast, OCB would have some

aspects of non-job-specific task proficiency, written and oral communication,

demonstrating effort, maintaining personal discipline, facilitating peer and team

performance, leadership, and administration. I would expect each citizenship behaviour

have varying amounts of these components and in some cases, none. For example,

voluntarily helping another team member may not involve written communication but it

may involve speaking to the other person. Alternatively, OCB that involves writing the

minutes for a volunteer meeting requires the skill of written communication. Notably,

the relationship between OCB and the performance components suggests the

involvement of interpersonal facilitation and motivational aspects. As such, antecedents

of these aspects are important to include when attempting to predict OCB.

Campbell’s (1990) performance model suggests that three performance

determinants can be used to describe all performance components, which are labelled:

declarative knowledge, procedural knowledge and skills, and motivation. Declarative

knowledge involves the understanding of factual information, rules, processes,

procedures, and principles that are related to the task. Procedural knowledge and skills

involves the capability that is created when the individual knows how to apply the

declarative knowledge to complete the task practically. Thirdly, motivation involves the

choice to perform the task. All three performance determinants are required to produce

the different types of performance components that exist within and across jobs.

Campbell (1990) expressed the relationship between a job performance

component and the three performance determinants in equation form. Campbell et al.

(1993) further extended Campbell’s earlier work to include the sub-aspects that would

be included in each of the performance determinants, which is summarised at Figure

3.1. Complex performance, such as OCB, can consist of different types of behaviours


within the same construct. The different types of behaviours are predicted by patterns of

performance determinants and performance components. The patterns of the

performance components would be expected to be different for each of the sub-

dimensions of OCB, despite having the same performance determinants. For example,

consider the two OCB sub-dimensions of sportsmanship and altruism. Sportsmanship

involves behaviour that avoids creating issues in the workplace, while altruism is an

active behaviour that helps other people or the organisation. Both behaviours may aim

to avoid conflict. However, the rationalisation for an individual to engage in

sportsmanship behaviour is to avoid the conflict that impacts on the game of work;

while altruism avoids conflict to help others. Subsequently, the same performance

determinants may be used with different combinations of performance components for

each type of behavioural sub-dimension. Importantly, Campbell et al. note that the

precise relationship between a performance component and the three performance

determinants would probably never be known, and most likely would vary across

different job performances.

x x PCi= ƒ [ Declarative

Knowledge

(DK)

Procedural

Knowledge

and Skills

(PKS)

Motivation

(M) ]

Where i = 1, 2, …k performance components

Facts

Principles

Goals

Self-knowledge

Cognitive Skill

Psychomotor Skill

Physical Skill

Self-Management Skill

Interpersonal Skill

Choice to perform

Level of Effort

Persistence of Effort

Figure 3.1

Equation representation of the job performance model (Campbell et al., 1993, p.43)


Campbell et al. (1993) suggest that declarative knowledge is an antecedent of

procedural knowledge and skill, as without the knowledge of the skill, the performer

cannot apply the knowledge to perform the behaviour. To consider the appropriateness

of their argument, take the case of a learner attempting to produce a behaviour, such as

shaving. If the learner has not gained all of the information and knowledge that is

required to complete the behaviour, such as using shaving cream instead of shaving

without a lubricant, the individual may complete part of the performance of shaving but

not necessarily at a competent level. As such, a competent performance requires the

knowledge to conduct the task, which the individual then applies in a practical setting.

Specifically, Campbell et al. argue that procedural knowledge and skills provide the

practical application of the declarative knowledge. However, the authors note that there

are occasions that the individual can produce a competent performance, such as social

behaviour, without fully understanding the performance. This may indicate that some

behaviours are produced through sub-consciousness processes, either with a biological

base or through learning that was not recognised by the learner. As such, Campbell et

al. suggest that declarative knowledge and practical knowledge and skills may have

some antecedents that are dispositional while others are learned. Both forms of these

antecedents should be considered when predicting performance.

The third performance determinant of motivation is also considered to have

dispositional antecedents. Campbell et al. (1993) argued this determinant is essential for

the production of performance. Unless the individual chooses to perform, works at an

appropriate level of effort, and persists at the necessary effort level, the performance is

not completed. The non-production of performance can therefore occur due to

motivational aspects, despite the individual possessing the knowledge and skills

required to produce the behaviour.


Both Campbell (1990) and Campbell et al. (1993) propose that declarative

knowledge and procedural knowledge and skills possess similar categories of

predictors. Table 3.2 summarises the predictors of the performance determinants. As

shown, these two performance determinants are broadly predicted by ability,

personality, interests, education, training, experience, and aptitude/treatment

interactions. Despite having the same predictor categories, declarative knowledge and

procedural knowledge and skills are different performance determinants. As such, they

are expected to have different aspects of the predictors for a performance component,

even if the categories are the same. To demonstrate this concept, once again consider

the behaviour of shaving. Competent shaving can involve the process of education to

develop declarative knowledge regarding the shaving equipment, lubricant, and the

shaving behaviour. Education can also enhance procedural knowledge and skills in the

areas of how to apply the lubricant, how to use the razor, and how to clean the razor.

While education is involved in both performance determinants, the aspects of learning

are different. As such, the predictor of education would vary between the two

performance determinants, despite being involved in both.

Table 3.2

Performance Determinant Predictors (Campbell, 1990)

Job Performance Determinants Predictors of Performance Determinants

Declarative Knowledge Ability, personality, interests

Education, training, experience

Aptitude/treatment interactions

Procedural Knowledge Ability, personality, interests

Education, training, practice, experience

Aptitude/treatment interactions

Motivation Independent variables as stipulated by various

motivation theories


In contrast to the stated predictors of declarative knowledge and procedural

knowledge and skills, Campbell (1990) did not specify motivation prediction, as the

predictors varied across different motivation theories. The lack of description of the

motivation predictors in both Campbell’s performance model and the work conducted

by Campbell et al. (1993), perhaps hints at the complexity of predicting performance

based on motivational aspects. Barrick, Stewart, and Piotrowski (2002) observed that

one of the reasons for the difficulty in identifying motivational predictors is that one

universally accepted framework to measure motivation has yet to be developed. As a

demonstration of the numerous theories that consider motivation, I have included a

summary of the major endogenous and exogenous motivation theories described by

Katzell and Thompson (1990) in Appendix B.

Motivation remains an important performance determinant to predict OCB,

especially given the discretionary nature of the performance (Organ et al., 2006). From

the many theoretical perspectives for motivation, Mitchell (1997) provides a relevant

definition which captures the necessary elements for predicting performance. He

describes work motivation as the “psychological processes involved with the arousal,

direction, intensity, and persistence of voluntary actions that are goal directed” (p.60).

Moreover, Kline (2003) suggests simply that motivation is the why the individual does

what he or she does.

While many consistent predictors had been identified for task performance (e.g.

see Doverspike et al., 2004), early job performance researchers who attempted to

identify motivationally-based predictors were not very successful. Notably, these

researchers tended to utilise task performance as their measure of performance, rather

than discretionary performance such as OCB. This methodology did not produce

consistent relationships between job performance and motivationally-based variables,


with a notable exception of the personality trait conscientiousness. While personality is

considered to be a motivationally-based variable, it is also described in Campbell’s

(1990) performance model as a predictor of declarative knowledge and procedural

knowledge and skill. Aspects of conscientiousness may have been predicting declarative

knowledge and procedural knowledge and skill that are required for task behaviour

rather than motivational aspects. As will be further discussed, the lack of consistently

observing a relationship between motivational variables and performance convinced

some researchers that there was no relationship. In contrast, other researchers continued

to explore the relationship and searched for other possible variables that when included

in the analysis may be able to demonstrate this relationship more consistently.

Through the efforts of researchers, such as Bateman and Organ (1983), Smith et

al. (1983), Campbell (1990), Borman and Motowidlo (1993), and, Campbell et al.

(1993), performance began to be considered as a multidimensional construct.

Importantly, the identification of a discretionary type of performance gave hope to

researchers investigating the relationship between motivation and performance.

Notably, Organ et al. (2006) stated that due to the discretionary nature of OCB, this type

of behaviour would be more related to motivationally-based variables than the

compulsory behaviours of task performance. Borman and Motowidlo (1993) also

suggested that task performance and OCB would have different antecedents. Task

performance would best be predicted by knowledge, skills and abilities, whereas the

antecedents for contextual performance (synonymous with OCB for the purpose of this

study) would involve dispositional factors.

In theorising the importance of motivation to predict OCB, Organ et al., (2006)

proposed there would probably be multiple and possibly overlapping motives that

would promote OCB in the workplace. As such, researchers have tended to investigate


the role of motivation indirectly through the measurement of specific attitudinal aspects.

This strategy enabled researchers to study motivation without requiring an investigation

of the whole construct. Notably, the pioneers of OCB antecedent research, including

Organ and his colleagues, focused on the job satisfaction-OCB relationship. As will be

discussed, job satisfaction is an attitudinal aspect that would affect the employee’s

motivation to produce discretionary behaviour in the workplace. The greater the level of

satisfaction, the more OCB would be expected to be produced (Organ, et al., 2006).

At the same time as the job satisfaction-OCB research was being conducted,

other researchers were considering the role of personality in the prediction of overall job

performance. Personality is a broad construct on which actions, thoughts, and feelings

are based. While motivation is the why, personality is considered the how a person acts,

thinks, and feels (Kline, 2003). From an Industrial-Organisational perspective,

researchers and practitioners have relied on the use of trait theory to describe

personality. Personality trait theory consists of two major aspects: firstly, personality

traits are relatively stable representing consistent behaviour over time and situation; and

secondly, personality traits influence work behaviour (Matthews, Deary, & Whiteman,

2009).

Over the past 30 years, the most common theoretical perspective utilised to

study the personality-performance relationship has been the five factor model of

personality (e.g. Barrick & Mount, 1991; Barrick, Mount & Strauss, 1993; Mount,

Barrick, & Stewart, 1998a; Tett, Jackson, & Rothstein, 1991; and Tett Jackson,

Rothstein, & Reddon, 1999). The five factor model, alternatively known as the Big

Five, provided researchers with a common and workable model of personality that was

particularly useful in the work environment. The five factors, which will be described

later in this chapter, are labelled openness to experience, conscientiousness,


extraversion, agreeableness, and neuroticism. It is important to note that not all

researchers have agreed with the usefulness of five factors and there have been

alternative models proposed that involved differing numbers of factors. For example,

some researchers have discussed the requirement for personality models be more

detailed and as such, should consist of more factors (e.g. Cattell, 1957: 16 factors;

Hogan, 1983: six dimensions of personality, which involved splitting the extraversion

factor into sociability and activity; Hough, 1992: nine factors of personality; and Hogan

& Holland, 2003: seven factor, which involved splitting the extraversion factor into

extraversion and ambition to form a larger factor of surgency, and splitting the openness

factor into two factors – intellect and interest in learning and achievement). In contrast,

other researchers have argued that three factors were sufficient (e.g. Eysenck, 1967,

1982).

Despite the alternative perspectives regarding the number of traits, the majority

of antecedent performance research has continued to utilise the five factor model.

McAdams (1992) recognised that the five factor model provided a sufficiently

comprehensive framework in the context of individual differences research. McCrae

and Costa (1991) argued that the five factor model was useful to provide a foundation

for systematically investigating personality and affect. It should also be noted that no

one theory of personality provides a total integrative theory of personality, but rather

each theory increases the overall knowledge of the personality construct (Feshbach &

Weiner, 1991). As such, personality theories aim to measure different aspects of the

larger and complex construct of personality and all theories have unique attributes and

criticisms. As no single theory provides a complete explanation of personality, it is

important to utilise the proven concepts in an eclectic manner that is appropriate for the

situation (Aiken, 1993). Many researchers propose that the five factor model provides a

sound and reliable model to describe personality traits in a workplace setting. As such,


the five factor model was selected as the framework on which to study personality traits

in the workplace in the current study.

Initial work by researchers using the five factor model compared measures of

performance which equated to task performance with the five personality factors, rather

than motivationally-based aspects of job performance. In terms of Campbell’s (1990)

performance model, personality was initially researched in conjunction with declarative

knowledge and procedural knowledge and skills, rather than the third determinant of

motivation. However, with the identification of OCB, researchers were able to study the

role the personality in the prediction of motivationally-based performance. Researchers

considering the personality-performance relationship could also investigate the

personality-OCB relationship. Moreover, these researchers could also integrate the

knowledge gained from broader OCB antecedent research, such as the relationship

between job satisfaction and OCB. When personality and job satisfaction were

investigated simultaneously, a stronger and more coherent theory was developed

regarding OCB antecedents (Organ et al., 2006).

While job satisfaction and personality have maintained key roles in the study of

OCB, they have lacked some consistency in their prediction results, as will be discussed

in this chapter. Given these findings, I argue that the prediction of OCB is more

complex than can be described by job satisfaction and personality alone. Additional

antecedent variables are therefore required to provide a more complete prediction

model. Notably, the attitudinal variable of affective disposition has been proposed as a

possible OCB antecedent and will be discussed in more detail later in this chapter.

Additionally, from a motivational perspective, OCB antecedent researchers have

investigated the construct of organisational commitment. Organisational commitment is

a global attitudinal variable that has been argued to be more stable than the attitudinal


variable of job satisfaction (Mowday, Steers, & Porter, 1979). As such, commitment

measures the individual’s attachment to the organisation, while job satisfaction involves

the individual’s satisfaction with specific aspects of the job conditions. Arnold, Cooper,

and Robertson (1998) suggested that commitment would lead to an increase of OCB,

particularly helping behaviours and increased conscientiousness.

Mowday et al. (1979) suggested that organisational commitment consisted of

two aspects: firstly, the individual’s identification with an organisation, and secondly,

the individual’s level of job involvement. Job involvement was considered as an

attitudinal aspect of job motivation and defined by Paullay, Alliger, and Stone-Romero

(1994) as the degree to which the individual is cognitively preoccupied and engaged in

the current job. While the relationship between job involvement and performance was

considered limited when the performance was measured as task performance (e.g.

Brown, 1996; Brown & Leigh, 1996); more recent studies have observed a relationship

between job involvement and OCB (e.g. Diefendorff, Brown, Kamin, & Lord, 2002,

which will be discussed in more detail in the personality section of this chapter;

Rotenberry & Moberg, 2007). Rotenberry and Moberg conceptualised their study of job

involvement and performance within the framework of social exchange theory. As

discussed previously, social exchange theory proposed that there was a psychological

exchange between the individual and the organisation/other employees. When applied

to the production of OCB, employees would exchange assistance with each other.

Rotenberry and Moberg suggested the reciprocation of the behaviours motivated

individuals to exchange additional OCB. This reciprocation of helping behaviour would

increase the level of job involvement of the individual in the current job, and

subsequently produce more OCB.


To investigate their theory, Rotenberry and Moberg (2007) collected data from

320 participants. The authors assessed three hypotheses, which proposed the existence

of positive correlations between job involvement, as measured by self-report, and the

supervisor scores on task performance, OCBI and OCBO. As described in Chapter 2,

OCBI were individually directed OCB, while OCBO was organisationally directed

OCB (Williams & Anderson, (1991). Importantly, the authors argued that these

correlations would exist beyond that produced through work centrality. Work centrality,

also known as work involvement, was argued to be a confounding variable that had

previously prevented job involvement from demonstrating correlations with

performance (e.g. Brown, 1996; Brown & Leigh, 1996). In contrast to job involvement,

which was considered as the level of involvement in the individual’s current job; work

centrality was formed from an historical perspective over the course of the individual’s

career and considered the individual’s value of work across all previous jobs.

Rotenberry and Moberg’s (2007) results showed positive correlations between

job involvement and work centrality (r=0.52, p<0.05), OCBI (r=0.32, p<0.05), OCBO

(r=0.13, p<0.01), and performance (r=0.15, p<0.01). Importantly, following regression

analysis, these correlations were observed for OCBI and in-role performance when

work centrality was controlled for. In contrast, OCBO did not demonstrate a significant

effect when work centrality was controlled for. Notably, Diefendorff et al., (2002) had

stated that work centrality was related to the OCB dimension of civic virtue, which was

a form of OCBO. As such, Rotenberry and Moberg suggested that job involvement and

work centrality were likely to be measuring similar aspects which were related to civic

virtue. However, given the results, job involvement provided a useful motivationally-

based construct on which to study the prediction of OCB.


Motivational aspects have been studied increasingly with OCB, with the

majority of OCB research exploring relationships between overall OCB or specific

dimensions of OCB (i.e. altruism and generalised compliance) and possible antecedents,

namely job satisfaction (e.g. Bateman & Organ, 1983; Smith et al., 1983). The increased

research effort has provided an opportunity to use meta-analytical techniques to

combine the collective results from multiple studies, thereby increasing the

generalisability of results. Additionally, the increased knowledge has enabled the use of

structural equation modeling to simultaneously assess the plausibility of theoretical

pathways between multiple variables. However, the majority of the research

methodology is correlational. Despite the correlational nature of the methodology, the

terminology for OCB research generally refers to the predictor variables as antecedents.

It is important to acknowledge that no statistical technique is able to provide evidence

of causality in non-experimental research (Kline, 2011). As such, caution must be taken

in the interpretation of correlational studies.

It is also important to recognise that the majority of OCB research has used

subjective measures. Subjective measures are based on judgments, feelings, and human

experience (Muckler & Seven, 1992). Scientific study has tended to support the use of

objective measures over subjective measures. One of the main reasons against using

subjective measures is due to reliability and validity issues. Aron and Aron (1999)

contend that subjective measures have been criticised for the lack of accuracy and

consistency of the rating, as subjective bias’ may influence the response. In contrast,

objective measures suggest that they are beyond subjective influences and present only

factual information. However, Muckler and Seven suggest that even objective measures

are interpreted subjectively. Moreover, in the study of abstract concepts, Muckler and

Seven argue that subjective measures are superior to objective measures due to the


difficulty of developing objectives items to measure these constructs. The resultant

difficulty exists as objective scales measure what is happening; as compared to

subjective scales that include not only what is happening, but also past experience,

present knowledge, intentions, and motivational levels. As such, subjective measures

can provide a richer and more unique measurement of human behaviour, including

OCB.

While subjective measures are considered superior over objective measures for

OCB research, it is important where possible to minimise potential bias that can

influence the analysis of data collected using subjective measures. Common method

variance is one such methodological issue that requires consideration to reduce bias that

was widely experienced with initial OCB research. Podsakoff, MacKenzie, Lee, and

Podsakoff (2003) noted that researchers have been concerned about common method

variance in behavioural research since the 1960s, as it was a major source of

measurement error that threatened the validity of a study’s results. While there were

statistical techniques to help reduce the effects of common method variance (see

Podsakoff et al., 2003), an important methodological issue when measuring OCB

involved incorporating multiple rater sources. Importantly, LePine et al. (2002) and

Allen, Barnard, Rush, and Russell (2000) suggested that different rater sources may

provide more effective ratings of the dimensions as they observed different types of

behaviour. For example peers may be better suited to observe behaviour interactions

between individuals, while supervisors may observe more organisationally directed

behaviours.

A common multiple observer method in the Industrial-Organisational

environment is 360 degree feedback, which is detailed in Chapter 4. This strategy was

used initially in the Human Resource domain as a professional developmental tool. The


360 degree feedback method enables an individual’s behaviour to be observed by

supervisors, peers, subordinates, and occasionally customers of the employee being

reviewed. The external feedback is then compared with the individual’s self-assessment

to enhance the employee’s knowledge of how they are perceived by others. For research

purposes, compiling observer feedback from a 360 degree perspective can provide a

more detailed picture of OCB. Organ et al. (2006) noted that further research was

required to investigate the effectiveness of multiple raters and suggested the use of 360

degree feedback for OCB research.

While researchers have advanced their methodology to more accurately predict

OCB, other researchers have remained convinced that ability should also be

investigated. Brown and Leigh (1996) suggested that motivational aspects have a role in

the prediction of performance; however, it would be equally likely that declarative

knowledge and procedural knowledge and skills, as described by Campbell (1990) and

Campbell et al. (1993), would have general knowledge and ability antecedents. Dawis

(2001) argued that ability provided the most useful prediction measures for performance

over other variables, including motivation. These abilities included general mental

ability (g), verbal intelligence, numerical intelligence, and spatial intelligence.

Moreover, Campbell’s (1990) model of performance identified that many aspects were

involved in the prediction of job performance, including: personality, interest,

education, training, experience, aptitude/treatment interactions, and motivation. These

aspects could be applied in different patterns across the eight factors of Campbell’s et

al. (1993) model. Consequently, while the initial emphasis for OCB research has been

on the motivational aspects of attitudinal variables, it is argued that predictor variables

that represent declarative knowledge and procedural knowledge and skills in the OCB

construct should also be identified and studied.


General mental ability has remained a consistent predictor of performance across

different jobs for declarative knowledge and procedural knowledge and skills.

Moreover, interpersonal skills were included as a procedural skill in Campbell’s et al.

(1993) performance model and have received increased research attention as an OCB

antecedent. As such, in addition to motivationally-based variables, consideration of the

ability contribution should be afforded to OCB antecedent research. Investigating

relevant OCB ability antecedents in conjunction with the traditional attitudinal and

dispositional variables could provide a more thorough picture of the complex construct

of OCB. It is also noted that some of the variables that contributed to motivational

prediction of OCB may also contribute to ability pathways, as described in Campbell’s

model. For example, personality would be expected to contribute to motivational

aspects to predict OCB and also ability pathways for procedural knowledge and skills.

The major antecedents used in OCB research efforts will now be discussed.

Job Satisfaction

The job satisfaction causes performance hypothesis is often referred to as the

‘Holy Grail’ of OCB investigations because the relationship seems obvious to the

layperson but has been difficult to prove through research (Weiss & Cropanzano, 1996).

Initially, job satisfaction was described as an affective state resulting from the

employee’s work experience. For example, Locke (1976) defined job satisfaction as a

“pleasurable or positive emotional state resulting from the appraisal of one’s job or job

experiences” (p. 1300). Over the past two decades, researchers (e.g. Fisher, 2000,

Weiss, 2002) have refined this position to argue that job satisfaction is an attitude,

consisting of an affective component and a cognitive component. The affective

component consists of emotions and feelings, while the cognitive component includes


beliefs, judgements, and evaluations based on the job or job situation. The layman’s

assumption that if employees are happy about their jobs, what they have to do, and/or

their colleagues, then they would be more likely to engage actively in task performance.

In other words: happy employees are productive employees.

Inconsistent results have plagued job satisfaction-performance research, which

in part has tended to fuel continued research. A debate has continued between those

researchers who suggested there was no actual relationship between job satisfaction and

performance and those who argued that a positive relationship did exist. An early

influential study that was recognised to have contributed to this debate was conducted

by Iaffaldano and Muchinsky (1985). These two researchers conducted a meta-analysis

of 74 empirical studies that were published in 70 articles, culminating in a sample size

of 12192. Their results identified a small positive correlation of 0.17 between job

satisfaction and performance. This result was similar to that identified by Vroom (1964)

who reported a median correlation of 0.14 when considering the findings of 20 studies.

However, despite actually finding supporting evidence that a small positive relationship

existed, the authors argued that the relationship between job satisfaction and

performance was practically non-existent because the correlation was too small. As

such, Iaffaldano and Muchinsky (1985) stated that while it was intuitive to relate these

two variables, no actual relationship existed.

While Iaffaldano and Muchinsky’s (1985) decision to reject the existence of a

small positive correlation between job satisfaction and performance could be debated, it

was equally important to consider why they produced a small correlation. Like all meta-

analytic studies, Iaffaldano and Muchinsky (1985) identified selection criteria for

studies and variables to be included or excluded from further analysis. Their selection

criteria for the performance variable involved the requirement to measure some form of


productivity. The authors also stated that studies were excluded if they contained

performance measures that considered aspects such as tardiness, absence, turnover, or

union grievances in order to gain a more singular construct of performance which would

aid interpretation of their results. This is particularly relevant for OCB research as these

latter types of removed behaviours were more likely to contain aspects of OCB-related

behaviour. Consequently, the authors focused on measuring task behaviour rather than

OCB. The results, from their perspective, therefore suggested that job satisfaction was

not practically related to task performance. However, their study did not specifically

assess the job satisfaction-OCB relationship.

In a more recent study Judge et al. (2001) published a meta-analytic study that

considered the results and conclusions of the Iaffaldano and Muchinsky (1985) study.

The authors argued that while Iaffaldano and Muchinsky (1985) provided advances in

comprehension and precision, their work contained several limitations. One suggested

limitation was the presence of publication bias through the exclusion of all unpublished

research. This type of bias can potentially ignore the results of articles, papers and

doctoral theses that were yet to be published. However, it could also be argued that

unpublished work should be excluded from the meta-analysis as they were not peer

reviewed and may lack the scientific research criteria. Judge et al. (2001) also

considered that Iaffaldano and Muchinsky underestimated the correlation between job

satisfaction and performance due to the statistical analytic techniques used. The most

important analysis that impacted on the results was how the average correlation between

job satisfaction and performance was produced. Specifically, Iaffaldano and Muchinsky

did not compute a correlation between a global measure of job satisfaction and

performance but rather averaged the correlations between each job satisfaction facet and


job performance. Judge et al. (2001) argued that this method downwardly biased the

mean correlation estimate.

In an attempt to strengthen the Iaffaldano and Muchinsky (1985) study, Judge et

al. (2001) utilised a methodology that included a range of studies. These studies were

taken from published articles and also included unpublished studies, data, conference

papers, technical reports, and doctoral theses that were written or published between the

period of 1967 to 1999. Additionally, for inclusion in their analysis, the authors required

a global measure of job satisfaction and a general measure of job performance, although

not specifically OCB. Notably, a review of the 312 studies that were included in the

meta-analysis identified only two titles specifically containing OCB related words;

extra-role and interpersonal evaluations. However, based on the formal naming of OCB

in 1983, it would be reasonable to assume that some of the overall performance

measures contained OCB-related items. Following the identification of OCB, this type

of behaviour would have been increasingly recognised in the workplace and

subsequently identified in day-to-day working activities.

Judge et al. (2001) achieved an estimate of the population correlation of 0.30

(k=312, N=54471; where k is the number of studies and N represents the total sample

size) for the job satisfaction-job performance relationship when corrected for

unreliability in the measures of job satisfaction and overall performance. Their result

was notably higher than the correlation achieved by Iaffaldano and Muchinsky (1985).

Moreover, utilising both qualitative and quantitative results from their study, Judge et

al. proposed an integrative model for the job satisfaction-performance relationship,

which is replicated at Figure 3.2.


The casual model at Figure 3.2 proposed by Judge et al. (2001) recognised the

complexity of job behaviour and incorporated many hypothesised and tested

relationships into a single unified framework. Their model was bidirectional and

included possible moderators and mediators. Importantly, the authors noted that further

research was required to test this model. While this may be the case, the model is too

complex to be tested in its entirety; only a few components can be tested at any one

time.

Moderators

Personality/self-

concept

Autonomy

Norms

Moral obligation

Cognitive accessibility

Aggregation

Level of analysis

Job

Satisfaction

Job

Performance

Mediators

Success and

achievement

Task specific

efficiency

Goal progress

Positive mood

Mediators

Behavioral

intentions

Low

performance

as withdrawal

Positive mood

Moderators

Performance-rewards

contingency

Job characteristics

Need for achievement

Work centrality

Aggregation

Figure 3.2

Integrative model of the job satisfaction-performance relationship. Replicated from Judge et

al. (2001, p. 390).


Importantly, the existence of a bidirectional job satisfaction-performance

relationship may not be observable consistently in both directions. For example, Riketta

(2008) conducted a small meta-analysis containing 16 studies that were written,

published, or presented between 1974 and 2006. The design of the study enabled a

causal examination of relationships between job attitudes and performance. The

methodology involved a panel design which was subsequently analysed using multiple

regression analysis. The inclusion criteria required job attitudes and performance to be

measured at a minimum of two separate occasions. Although the study was not

specifically aimed to review OCB, the performance items were coded for both in-role

(traditional job behaviour) and extra-role (synonymous with OCB) behaviour. Five

studies focused on extra-role behaviour and 11 studies considered in-role behaviour.

The study analysed data between performance and job attitudes, including job

satisfaction, on single occasions and compared the results over a time lag, with the

average time lag of nine months. Job attitudes were found to be predictors of overall job

performance, albeit small (β=0.06). The effect was stronger for the commitment to

performance pathway (β=0.08) than for the job satisfaction to performance pathway

(β=0.03). Riketta (2008) indicated that the effect was stronger for shorter time lags (1–6

months, β=0.12) than the longer time lags (7-12 months; 13+ months; both ns).

Conversely, the only statistically significant finding for the effects of performance on

job attitudes was a small negative effect for a moderate time lag (β=-0.08, 7-12 months).

Subsequently, the results suggested that job attitudes, including satisfaction, were more

likely to influence performance than performance was to influence attitudes. While this

study provided important causal relationship results, further research was required due

the limitations of the size of the meta-analysis and the use of self-report measures for

the extra-role performance.


Riketta’s (2008) study supported a uni-directional relationship to predict OCB

from job satisfaction and job commitment, however no relationship was observed for

the reverse direction of the relationship. While a bi-directional relationship was not

observed in Riketta’s study, a bi-directional relationship could be expected between

OCB and attitudes due to motivation. Notably, Brown (1996) argued that motivation

was likely to be an antecedent and consequence of the attitude of job involvement. If

this conjecture is considered within the social exchange theory framework, OCB would

be produced as part of a psychological exchange between the individual and the

organisation/other employees, as previously discussed by Rotenberry and Moberg

(2007) and (Organ et al., 2006). Moreover, the reciprocity of the exchange between

employees and the workplace and employees would motivate individuals to produce

more OCB. The increased motivation would in turn increase job involvement and job

satisfaction, which has been observed to increase OCB (e.g. Riketta, 2008). The cycle of

reciprocation would suggest a bi-directional relationship. However, given the

complexity of work behaviour, it is possible that there are many situational variables

that could influence the individual’s level of job satisfaction. Notably, the model

produced by Judge et al. (2001) hypothesised numerous moderators and mediators to

the job satisfaction-performance relationship. Given this complexity, the production of

OCB in the workplace may not produce a significantly strong reverse affect to

consistently observe an increase in job satisfaction. In contrast, research has been more

successful when the prediction of OCB is studied through a uni-directional pathway

from attitudes to OCB.

While early studies considered the job satisfaction-overall performance

relationship, a meta-analytic study that considered the specific relationship of job

satisfaction-OCB was carried out by Organ and Ryan (1995). The importance of their


study was that it investigated this relationship, while also including other attitudinal

predictors and moderating variables, including organisational citizenship. In their study

Organ and Ryan conducted a quantitative review of 55 studies that were published

between 1983 and 1994. Interestingly, only two studies in this analysis were also

utilised in the meta-analysis conducted by Judge et al. (2001). The majority of the

included studies utilised the OCB measure developed by Smith et al. (1983), while job

satisfaction was measured using overall or global measures of job satisfaction. Organ

and Ryan examined the job satisfaction-OCB relationship through investigating

multiple variables to predict OCB. Specifically, they considered the dispositional

variables of personality and affective disposition, the motivational aspect of

organisational citizenship, other potential moderators, and demographic information.

Personality was measured using the traits of agreeableness and conscientiousness

(McCrae & Costa, 1987), while affective disposition was assessed using positive and

negative affectivity (Watson & Clark, 1984). Organisational commitment was measured

using an overall score, and also separated into two types – affective and continuance.

The other potential moderators included fairness and leader supportiveness, while the

authors also collected the demographic information of tenure and gender. Similar to

Borman and Motowidlo (1993), the authors argued that different aspects of performance

would have different antecedents. Specifically, they hypothesised stronger relationships

between job attitudes/disposition and OCB, than between job satisfaction and overall

job performance/traditional core behaviours that had been identified in previous studies

(e.g. Iaffaldano & Muchinsky, 1985; Vroom, 1964).

Results from Organ and Ryan’s (1995) study indicated that job satisfaction was

moderately correlated with OCB for non-managerial and non-professional employees

(ρOCB= 0.28, N= 2845; where ρ is the estimated population true-score correlation which


was corrected for unreliability and N is the sum of all the samples in the analysis).

Additionally, the authors observed small correlations for both altruism and generalised

compliance with overall commitment (ρA= 0.25, ρGC= 0.23) and affective commitment

(ρA= 0.32, ρGC= 0.30). However, no significant correlations were observed between the

two factors of OCB and the continuance score of commitment. Organ and Ryan

concluded that meaningful correlations were not generally found between the OCB

factors of altruism and generalised compliance and the following variables: personality

(agreeableness: ρA=0.13, N=916; ρGC=0.11, N=916) or affective disposition (positive

affect: ρA=0.08, N=869, ρGC=0.07, N=934; and negative affect: ρA=.-0.06, N=1201,

ρGC=-0.12, N=847). The notable exception to this pattern was the existence of a

moderate correlation between conscientiousness and generalised compliance (ρGC=0.23,

N=1231). Organ and Ryan suggested that while dispositions could provide orientations

to give and receive attitudinal related behaviour, it was more likely that these

contributions were indirect, with the exception of conscientiousness. Moreover, as the

correlations between OCB and the variables of job satisfaction, organisational

commitment, and fairness, were of similar size, Organ and Ryan suggested that no

single attitudinal measure was superior. Alternatively, as these variables all possessed

strong inter-correlations, perhaps they were measuring an ‘m’ factor, which could be

akin to morale. This argument supported the idea of motivational influences, such as job

involvement, that were measured through the attitudinal antecedents of OCB.

The correlation between job satisfaction and OCB that was observed by Organ

and Ryan (1995) was higher than that found in the previous meta-analysis conducted by

Iaffaldano and Muchinsky (1985). Notably, their results were similar to the correlation

demonstrated by Judge et al. (2001) and Riketta (2008). Based on the combination of

their results and those of Iaffaldano and Muchinsky (1985), Organ and Ryan (1995)


suggested that job satisfaction was related more to OCB-type job performance rather

than task performance behaviours.

From the previous discussion it would be reasonable to consider that the job

satisfaction-performance relationship, and more specifically the job satisfaction-OCB

relationship, was finally gaining reasonable support. However, the contentious nature of

this relationship continued to interest researchers on both sides of the debate.

Importantly, while some researchers remained convinced of the presence of the job

satisfaction-performance relationship, others continued to argue against the existence of

the relationship. In a recent meta-analysis, Bowling (2007) argued that the job

satisfaction-performance relationship was spurious once the five factor model

personality traits were statistically controlled for. His study utilised previously

published meta-analyses and also published and unpublished studies to investigate

relationships between organisation-based self-esteem, work locus of control, job

satisfaction, and job performance. Notably, the job satisfaction-performance

relationships produced small but statistically significant correlations ranging between

0.09 and 0.23. Bowling argued that the correlations were too small to be practically

significant, similar to the argument put forth by Iaffaldano and Muchinsky (1985).

While his conclusion can also be debated, as even small associations could be

practically significant to the organisation’s bottom line (Burns, 1997), it is important to

note that Bowling’s methodology may have reduced the size of the actual correlation

between job satisfaction and OCB. Similar to Iaffaldano and Muchinsky (1985), he used

an overall performance measure rather than OCB and subsequently did not assess the

relationship between job satisfaction and OCB. In fact, Bowling suggested that job

satisfaction did not cause performance, however, when employees were satisfied, there

may be other direct benefits to the organisation. Specifically, he noted that other


researchers have demonstrated that job satisfaction had relationships with specific types

of behaviours, including: organizational citizenship behaviours, counterproductive work

behaviours, turnover intention and actual turnover, and employee lateness.

While the initial research of the job satisfaction-performance relationship has

suggested that job satisfaction did have a role to play in the prediction of OCB, the

results have been inconsistent and small to moderate at best. Researchers have also

begun to consider other additional antecedents, including personality. Investigating

multiple OCB antecedents simultaneously provided a greater level of complexity to the

prediction of OCB. The next section details the study of personality as an antecedent of

overall job performance and OCB.

Personality

The study of personality is one of psychology’s oldest sub-fields and dates back

to 400 B.C., when Hippocrates considered there were four basic types of personality,

which correlated with four bodily substances known as humors (Atkinson, Atkinson,

Smith & Bem, 1993). With over 2400 years of attention it is not surprising that

personality has remained an important construct for researchers and practitioners alike.

Although there are many fundamental theoretical variations, Pervin (1990) considered

theorists generally agreed that personality was extremely complex and the development

of different theories simply emphasised this complexity and its importance.

There are five main principles that are generally accepted across the various

theories of personality. These principles are: personality was an organised whole;

personality was organised in patterns which were observable and measurable;

personality had a biological basis which was developed through social and cultural


environments; personality had both superficial and deeper core components; and

personality had both common and unique components (Ivancevich, Olekalns, &

Matteson, 1997). Consistent with these five principles, Maddi (2001) defined

personality as:

“a stable set of tendencies and characteristics that determine those commonalities and

differences in people’s psychological behavior (thoughts, feelings, and actions) that

have continuity in time and they may not be easily understood as the sole result of the

social and biological pressures of the moment”( p. 8).

While the majority of personality research had investigated personality

psychopathology, over the past century there has been an increasing effort to understand

normal personality in the Industrial-Organisational context. One of the theories

commonly utilised in Industrial-Organisational research is trait theory. Trait theory

suggests that an individual's personality is defined by traits and these traits are enduring

characteristics that can describe an individual’s behaviour (Robbins, Waters-Marsh,

Cacioppe, & Millet, 1994). As such, trait theory can be seen as a description of the

individual’s personality rather than a theory in its own right. Matthews et al. (2009)

suggested that while traits could not be expected to fully explain an individual’s

behaviour, they could provide a coherent basis on which to build personality theory.

Allport (1961) argued that personality traits were inherited and then shaped

through our experiences. Eysenck (e.g. 1967; 1982; and 1998, as originally cited in

1947) had also argued that traits were genetically based. He considered that the genetic

makeup enabled specific ‘conditioning’ thus resulting in variation of personality

between individuals. Similarly, Maddi (2001) suggest that the study of personality

involved comparing the similarities and differences between people, which were either

inherent or learnt. The inherent/inborn similarities and differences were considered the

core of personality. In contrast, if the similarities and differences were learned, Maddi

(2001) suggested they represented the periphery of personality. The core tendencies and


characteristics individuals possessed at birth would only change due to maturation. In

contrast, peripheral characteristics could be influenced by external factors, including

culture, family, and situation. As the perspectives from Eysenck (1998) and Maddi

(2001) were very similar and highly respected in trait theory, Figure 3.3 provides a

diagrammatic representation combining their hierarchical personality structures.

Figure 3.3 describes a four-level structure of personality. The fourth level is

considered to be the broadest and most influential level of personality, which is

Core

Statement

Development

Statement

Periphery

Statement

Data

Statement

Type Level

Trait Level

Habitual

Response Level

Specific

Response Level

Maddie’s

Personality

Structure (2001)

Eysenck’s

Personality

Structure (1998)

Figure 3.3

Four level schematic personality structure based on the work of Eysenck (1998) and Maddi

(2001).


represented by the personality type or core statement of the individual. Maddi (2001)

argued that this level corresponded to the overall direction of living and consisted of

goals, ideals, and instincts. The trait level forms the third level of personality and is also

considered a dominant influence on the individual’s personality, albeit less so than the

fourth level. This level is characterised by the existence of rigid thoughts, feelings and

actions by the individual. The rigidity occurs despite the individual having an

interaction with the broad influences of family, work, society, and culture. These two

broader levels of personality influence the more specific personality levels. This

influence would enable the production of responses across a range of situations within

the second level of personality. Maddi (2001) describes the second level as the

periphery statement. The behaviours identified at this level are generally characteristic

of the individual and occur across various situations (Eysenck, 1998).

The most specific level of personality is the first level of personality (Eysenck,

1998; and Maddie, 2001), which is situationally-based. This level is labelled the specific

response level or data statement and occurs when the individual has thoughts, feelings

and actions that exist in everyday circumstances. These thoughts, feelings or actions

may be observed once but are not necessarily repeated. In other words, this type of

behaviour can change from situation to situation. As the overall structure shows, the

broadest levels can influence the lower and more specific personality levels to produce

behaviour that is consistent with the individual’s personality. Notably, the broader

structure can only influence, rather than guarantee the behaviour of the individual. This

suggests that individuals can produce behaviour that is inconsistent with the broader

personality structure in some situations.

In the personality-performance context, specific personality traits were

researched to determine their relationship, if any, with job performance. Personality trait


measurement to predict job performance dated back to approximately 1909 when

Heymans and Wiersma studied the trait ratings of 400 physicians for over 2,500

individuals (Tupes & Christal, 1961; reprinted 1992). While there was an initial surge in

the first half of last century to use of personality to predict job performance, inconsistent

results and the intrusive nature of testing created public distrust due to fears of misuse

(Kaplan & Saccuzzo, 1997). However, as noted by Barrick and Mount (2005), by the

1980s there was increasing support that personality could predict job performance.

Similar to the job satisfaction-performance relationship, the personality-

performance relationship was also based in part on a ‘common sense’ foundation and

has continued to interest researchers and practitioners alike. Notably, Barrick and

Mount (2005) contended that this interest remained worthwhile because all workers

have a personality. Moreover, managers did not necessarily require peer reviewed

research to believe that there may be a right or wrong personality for employees. With

the choice of applicants, a manager would be more likely to hire a potential employee

who was dedicated, assertive, open to change, altruistic and confident over an applicant

who demonstrated irresponsibility, untrustworthiness, hostility, non-team contributor, or

inflexibility. The major reason for the returned interest in personality trait theory, as

discussed earlier, was due to the development of the five factor model of personality

(Barrick & Mount, 2005).

Five Factor Model of Personality

The five factor model of personality had been useful because it was a widely

supported framework that enabled a consistent approach to identify and assess the

personality-job performance relationship. Importantly the model did not claim to

provide a complete theory of personality, but rather a broad description of personality


traits. Nevertheless, the five broad dimensions of personality have demonstrated

robustness (Digman, 1996). This personality model has increased the predictability of

the complex nature of human behaviour at work through the consideration of specific

personality traits. McCrae and John (1992) argued that the “long history, cross-cultural

replication, and empirical validation across many methods and instruments makes the

five factor model a basic discovery of personality psychology – core knowledge upon

which other findings can be built” (p.207). Moreover, McCrae and John contended that

the five factor model has demonstrated its effectiveness not only in the Industrial-

Organisational context but has also been successfully applied to the domains of

educational psychology, forensic psychology, and health psychology.

The original five factor model was produced by Tupes and Christal (1961;

reprinted 1992) and later replicated by Norman (1963). Both of these studies utilised the

work conducted by Allport and Odbert (1936) and Cattell (1957). Allport and Odbert

had identified a list of approximately 4,500 trait terms by comparing around 18,000

trait-like words from an unabridged English dictionary and then eliminated close

synonyms and obscure words. Allport (1961) separated these traits into two categories,

firstly common traits – shared cultural/societal traits, and secondly individual traits –

unique to the individual. Cattell (1957) increased Allport and Odbert’s list of traits

using psychological and psychiatric literature and then reduced the trait terms to 171.

Cattell asked college students to rate their friends using 171 trait terms. This

information was factor analysed and reduced to 16 basic dimensions, which were

considered source traits and led to the development of the 16PF Questionnaire.

However, using Cattell’s research traits, Tupes and Christal (1961; reprinted 1992) and

Norman (1963) were unable to reproduce the 16 dimensions, both finding that five

factors accounted sufficiently for the data.


Trait theory was the basis of development for the five-factor model and also

enabled personality to be measured and described in the workplace. This was achieved

through the development of a five-factor model self-report inventory to measure the five

broad dimensions, known as the NEO PI (McCrae & Costa, 1985). Neuroticism,

extraversion, and openness to experience were assessed using facet scales; while

agreeableness and conscientiousness were assessed initially with global measures. Facet

scales for agreeableness and conscientiousness were developed two years later (McCrae

& Costa, 1987).

Costa and McCrae (1992a) consolidated their previous work culminating in the

development of the Revised NEO Personality Inventory (NEO PI-R), which was still

used extensively in personality research, particularly in organisational settings. The

NEO PI-R measured the five personality dimensions in order to impart an overview of

the individual, whilst also analysing six facets in each dimension to provide a more

thorough investigation. The five factors represented the core tendency of the individual

as described by Eysenck (1998) and Maddi (2001). The six facets represented the trait

level and combined to form a personality factor. These facets were determined using a

top-down factor analytic approach in order to provide representative and statistically

repeatable facets (Costa & McCrae, 1995). Interestingly, in comparison to most research

populations in the study of personality that use typically mentally ill norming samples,

Costa and McCrae (1992a) used a normal working population to develop and assess

their measure of the five factor model. A summary of Costa and McCrae’s (1992a) five

factor model with the facet scales is included at Table 3.3. A detailed description of the

NEO-PI-R facets scales, as adapted from Costa and McCrae (1992a), is contained in

Appendix C.


Table 3.3

NEO-PI-R Measure of the Five Factor Model (Costa & McCrae, 1992a)

Personality Dimensions Facets Measured for the Dimension

Neuroticism Anxiety, Angry Hostility, Depression, Self-

Consciousness, Impulsiveness, Vulnerability

Extraversion Warmth, Gregariousness, Assertiveness, Activity,

Excitement Seeking, Positive Emotions

Openness to Experience Fantasy, Aesthetics, Feeling, Actions, Ideas, Values

Agreeableness Trust, Straightforwardness, Altruism, Compliance,

Modesty, Tendermindedness

Conscientiousness Competence, Order, Dutifulness, Achievement Striving,

Self-Discipline, Deliberation

The first factor of personality has been termed neuroticism, which represented

the emotional dimension of personality (McCrae & Costa, 1987). This dimension was

the tendency to experience or feel negative emotions such as anxiety, depression, anger,

and embarrassment. As shown at Table 3.3, Costa and McCrae (1992a) measured the

neuroticism dimension using the facets of: 1. anxiety – the tendency to be apprehensive,

fearful, prone to worry, nervous, tense and jittery; 2. angry hostility – the capacity to

experience anger and related states such as frustration and bitterness; 3. depression – the

propensity to experience depressive affect; 4. self-consciousness – the inclination to be

uncomfortable around others, sensitive to ridicule and prone to feelings of inferiority; 5.

impulsiveness – the lack of ability to control cravings and urges; and 6. vulnerability –

the inability to cope with stress. High scores on this dimension corresponded to being

tense, anxious, nervous, moody, worrying, and touchy, while low scores included

descriptions of stable, calm, contented, and unemotional (John, 1990).

The personality factor extraversion measured the social dimension of personality

and contained the elements of sociability, cheerfulness, activity level, assertiveness, and

sensation seeking (McCrae & Costa, 1987). The NEO-PI-R measured extraversion


using the facets of: 1. warmth – issues of interpersonal intimacy; 2. gregariousness – the

preference for other people's company; 3. assertiveness – the capacity to be dominant,

forceful and socially ascendant; 4. activity – the ability to be rapid in tempo and

vigorous movement, a sense of energy, and a need to keep busy; 5. excitement-seeking

– the craving for excitement and stimulation, and enjoy bright colours and noisy

environments; and 6. positive emotions – the tendency to experience positive emotions

such as joy, happiness, love and excitement (Costa & McCrae, 1992a). Individuals with

high scores on extraversion were considered talkative, assertive, active, energetic,

outgoing, outspoken, and dominant, while low scores described quiet, reserved, shy,

silent, withdrawn, and retiring (John, 1990).

The third factor of the five factor model of personality was openness to

experience and represented the broad domain of intellect because it considered

openness, culture, and intellect (Digman, 1989). Openness to experience contained the

facets of: 1. fantasy – the capacity to be open to fantasy and have a vivid imagination

and an active fantasy life; 2. aesthetics – a deep appreciation for art and beauty; 3.

feelings – the ability to be receptive to one's own inner feelings and emotions and the

evaluation of emotion as an important part of life; 4. actions – a willingness to try

different activities, go to new places or eat unusual foods; 5. ideas – an intellectual

curiosity that allows an active pursuit of intellectual interests for their own sake, and an

open-mindedness and a willingness to consider new or perhaps unconventional ideas;

and 6. values – to possess the readiness to re-examine social, political and religious

values (Costa & McCrae, 1992a). High scores in this dimension were described as wide

interest, imaginative, intelligent, original, insightful, curious, and sophisticated, while

low scores tended to indicate commonplace, narrow interests, simple, shallow, and

unintelligent (John, 1990). McCrae and Costa (1987) made the important distinction


between openness and intelligence, because although an open individual was generally

perceived by themself and others to be more intelligent, the two have been shown to be

separate dimensions of individual differences. The moderate correlations may be

explained through intelligence predisposing individuals to openness or alternatively,

openness develops intelligence.

Agreeableness considered the moral dimension of personality (McCrae & Costa,

1987). Costa and McCrae (1992a) measured the agreeableness dimension using the

facets of: 1. trust – the ability to believe others are honest and well-intentioned; 2.

straightforwardness – the capacity to be frank, sincere and ingenuous; 3. altruism – the

tendency to have an active concern for others' welfare as shown in generosity,

consideration of others and a willingness to assist others in need of help; 4. compliance

– characteristic reactions to interpersonal conflict; 5. modesty – the potential to be

humble and self-effacing but are not necessarily lacking in self-confidence or self-

esteem; and 6. tender-mindedness – the attitudinal ability for sympathy and concern for

others. High scores on this factor were described as sympathetic, kind, appreciative,

affectionate, soft-hearted, warm, and generous, while low scores corresponded to fault-

finding, cold, unfriendly, quarrelsome, hard-hearted, and unkind (John, 1990). McCrae

and Costa (1987) indicated that extreme scores on agreeableness could represent

maladaptive behaviour.

Named Conscientiousness, the final factor of the five factor model represented

the organisational dimension of personality. Conscientiousness was measured using the

facets of: 1. competence – the sense one is capable, sensible, prudent and effective; 2.

order – the preference to be tidy, well organised and keep things in their proper order; 3.

dutifulness – to be governed by conscience; 4. achievement striving – the tendency to

have high aspiration levels, work hard to achieve goals, diligent, purposeful, and have a


sense of direction in life; 5. self-discipline – the ability to begin tasks and carry them

through to completion despite boredom and other distractions; and 6. deliberation – the

capacity to think carefully before acting (Costa & McCrae, 1992a). High scores stood

for organised, thoroughness, planfulness, effectiveness, and responsibility, while low

scores were described as careless, disorderly, frivolous, irresponsible, slipshod,

undependable, and forgetfulness (John, 1990).

Five Factor Model of Personality and Job Performance

The five factor model has been widely utilised to explore the personality-job

performance relationship. A key meta-analytic study investigating five factor model and

how the dimensions were related to job performance was conducted by Barrick and

Mount (1991). Specifically, their study considered each of the five personality

dimensions against three performance criteria: job proficiency, training proficiency, and

personnel data. Based on the data of 117 studies across five occupational groups

(professional/police/managers/sales/semi-skilled) and conducted between the years of

1952 and 1988, results indicated there was a relationship between personality and job

performance. However, this relationship varied across occupations and personality

dimensions. Specifically, conscientiousness predicted performance consistently across

all five occupations with small correlations ranging from 0.20 to 0.23. Extraversion was

related slightly to performance in sales (r=0.15) and managerial jobs (r=0.18), which

were considered the two occupations involving the most interpersonal skills

requirement. Results also indicated that openness to experience and extraversion were

able to predict training proficiency across various occupations. In contrast, the

dimension of agreeableness was not significantly correlated with job performance, even

when interpersonal skills were required.


While the meta-analysis conducted by Barrick and Mount (1991) was often

cited, the personality-performance correlations were small to moderate at best. In a later

meta-analysis conducted in the same year, Tett et al. (1991) analysed 97 studies and the

results generally supported those found by Barrick and Mount (1991). The corrected

correlation between overall personality and job performance was 0.24, job performance

correlations to the five dimensions of personality were: neuroticism (r=-0.22),

extraversion (r=0.16), openness (r=0.27), agreeableness (r=0.33), and conscientiousness

(r=0.18). Interestingly, Tett et al. observed that agreeableness held a moderate

correlation with job performance, which contrasted with the non-significant result in the

study conducted by Barrick and Mount (1991). Notably, a further meta-analysis

conducted by Mount et al. (1998a) produced results that supported the meta-analysis

conducted by Tett et al. (1991). Mount et al. (1998a) reviewed 11 studies to investigate

what personality dimensions of the five factor model were associated with jobs that

involved interpersonal relationships. While this study will be discussed in more detail in

the interpersonal skills section of this chapter, it is important to note that their results

indicated that conscientiousness, agreeableness, and emotional stability were all

positively related to jobs that involved interpersonal interactions (r=0.26; r=0.21; and

r=0.18, respectively).

A later study conducted by Barrick et al. (2002) also produced relationships

between personality traits and performance. The authors assessed the mediating effect

of cognitive-motivational work orientations on the relationship between personality

traits and sales performance. Their study collected data from 164 telemarketing sales

representatives. The results indicated that striving for status and accomplishment work

orientations mediated the effects of extraversion and conscientiousness on sales

performance. This suggested that extraversion and conscientiousness had indirect


pathways to the prediction of performance. As such, motivational aspects were

important to consider as additional variables to be included within an OCB prediction

model. However, while the mediation of the extraversion-performance relationship was

considered relatively straightforward to interpret through status striving, Barrick et al.

argued that the conscientiousness-performance relationship was more complex.

Specifically, the authors argued that conscientiousness had direct influence on

performance, while also being mediated through achievement striving and status

striving. As such, different facets of conscientiousness probably related to different

work attitudes and motivations. Additionally, while agreeableness possessed a weak

correlation with the communion striving work orientation (r=0.15), neither

agreeableness nor communion striving were related to sales performance. Once again,

agreeableness did not demonstrate a strong role in the prediction of performance within

a motivational context.

Five Factor Model of Personality and Organisational Citizenship Behaviour

Similar to the job satisfaction-performance research, over time some

personality-performance researchers have shifted their focus from overall job

performance to specific types of performance. However, as was discussed earlier in this

chapter, this study methodology found minimal support for direct relationships between

personality and dimensions of OCB, with the exception of conscientiousness (e.g.

Organ & Ryan, 1995). Notably this finding had not reduced the research effort to

investigate this relationship. Researchers and theorists continued to argue the

importance of the role of personality in the prediction of OCB (e.g. Barrick & Mount,

2005; and Organ et al., 2006).


Consistent with the hierarchical personality structure as described by Maddi

(2001) and Eysenck (1998), personality traits could influence the individual’s behaviour

in the workplace. Notably, this behaviour could be in the form of OCB. Recent studies

exploring the personality-OCB relationship have supported this argument finding

correlations between these two variables. In their study to examine the ability of job

involvement and work centrality to predict OCB and task performance, Diefendorff et

al. (2002) also considered the personality traits of conscientiousness and agreeableness.

Similar to the study conducted by Rotenberry and Moberg (2007), the authors argued

that job involvement and work centrality should be viewed as a state versus trait

construct. Data were collected from 130 employed undergraduate students and results

indicated that job involvement was positively correlated with the OCB dimensions of

altruism (r=0.20), civic virtue (r=0.34), sportsmanship (r=0.18), and conscientiousness

(r=0.20), but not courtesy. In contrast, work centrality was only correlated with civic

virtue (r=0.19). This indicated that job involvement was a better predictor of OCB than

work centrality. The personality traits of conscientiousness and agreeableness were

correlated with job involvement (r=0.34 for both traits), while only conscientiousness

was correlated with work centrality (r=0.18). In contrast to other studies,

conscientiousness was not observed to be significantly correlated with OCB, while

agreeableness was correlated only with sportsmanship (r=0.26). Neither personality

trait was significantly correlated with in-role performance. The main contribution of this

study was the findings that job involvement was related to personality traits and

appeared to be a valid predictor of OCB. However, the study utilised a student

population and the results were limited in the generalisability. As such further research

was required to consider the examined relationships, particularly with reference to

personality traits and OCB.


In a more recent study, Singh and Singh (2009) investigated OCB and its

relationship with four of the five personality traits of the five-factor model. The first

three hypotheses involved the ability of the individual to be interpersonally competent,

while the fourth hypothesis was concerned with the individual being predisposed to

perform extra tasks in the workplace. Specifically, their first hypothesis was that

neuroticism would be negatively related to OCB, as individuals who were high in

neuroticism expressed a lot of emotion. As discussed earlier in this chapter, neuroticism

related emotions include: anxiousness, depression, angriness, embarrassment, worry,

and insecurity. These emotions are likely to be more prevalent during times of stress or

in difficult situations and thus inhibit social behaviour (Singh & Singh, 2009). The

second hypothesis was that extraversion would be positively associated with OCB, as

this trait was a key disposition of social behaviour. Individuals who scored highly on the

extraversion trait were more sociable, assertive, active, bold, energetic and expressive.

The authors argued that highly extraverted individuals were more likely to display OCB

as they would produce more flexible behaviour. The third hypothesis proposed by Singh

and Singh was that agreeableness would be positively related to OCB, as being

agreeable would enhance interpersonal performance. Individuals who scored highly on

agreeableness were flexible and generally friendly, helpful, cooperative, courteous, and

good natured. As such, individuals high in agreeableness would interact with a higher

level of interpersonal competence and be more likely to produce OCB, than individuals

with low scores on agreeableness. The final hypothesis was that individuals who scored

highly on conscientiousness would produce more OCB. Individuals who scored highly

on conscientiousness were generally described as neat, punctual, careful, self-

disciplined, and reliable. As such, the authors argued that these individuals would be

predisposed to extend their performance beyond the required task performance to also

produce more discretionary behaviours, including OCB.


To assess their four hypotheses, Singh and Singh (2009) collected data from a

sample of 188 front level male managers who worked in either public or private Indian

organisations. The data included demographic information, OCB, and personality traits.

The authors used the OCB scale developed by Podsakoff, MacKenzie, Moorman, and

Fetter (1990) and measured the OCB dimensions of altruism, conscientiousness, civic

virtue, courtesy, and sportsmanship. This OCB scale was analysed with data collected

from the NEO Five Factor Inventory (NEO FFI). The NEO FFI is a shorter form of the

NEO PI-R and was also developed by Costa and McCrae (1992b). Support was

observed to varying degrees for all four hypotheses. Consistent with previous studies

(e.g. Barrick & Mount, 1991; Organ & Ryan, 1995), results indicated that the strongest

and most consistent predictor across all of the dimensions of OCB was the personality

factor of conscientiousness with positive correlations ranging from 0.25 to 0.48.

Extraversion was also positively related to all factors of OCB with relationships ranging

from 0.26 to 0.37, while agreeableness was strongly related to all dimensions with the

exception of civic virtue (correlations ranging from 0.26 to 0.46). In contrast,

neuroticism was negatively correlated with sportsmanship (r=-0.24), courtesy (r=-0.24),

and altruism (r=-0.16), but unrelated to the OCB dimensions of conscientiousness or

civic virtue. Their results suggested that the personality traits of conscientiousness and

extraversion had the strongest roles in the prediction of OCB, while agreeableness and

neuroticism had more limited roles.

Notably Singh and Singh (2009) did not measure openness to experience.

Openness to experience is generally not studied in the prediction of OCB. This factor

had demonstrated a relationship with success in learning and training (e.g. Barrick &

Mount, 1991; Mount & Barrick, 1998) rather than in the prediction of job performance.

In fact, Organ et al. (2006) could not identify a theoretical reason for the inclusion of


openness of experience in the prediction of OCB. However, Organ et al. contended that

people who were active, outgoing, and enjoyed being around other people

(extraversion) or friendly, good natured, and co-operative with others (agreeableness),

or organised, trustworthy, and persistent (conscientiousness) would be more likely to

engage in OCB type behaviours.

Summary of Job Satisfaction, Personality, and Organisational Citizenship

Behaviour

Job satisfaction was the original antecedent investigated to predict OCB (e.g.

Organ, 1777; Bateman & Organ, 1983; Smith et al., 1983). These early studies observed

small positive correlations between job satisfaction and OCB. In contrast, personality

was studied initially as an antecedent for overall performance (e.g. Barrick & Mount,

1991). It is noted that these earlier performance studies were actually measuring the

traditional form of performance, namely task performance. However, as researchers

have considered the possibility that job performance was a multi-dimensional construct,

personality has also been investigated as an antecedent for OCB (e.g. Organ & Ryan,

1995; Singh & Singh, 2009). Interestingly, some aspects of personality, for example

conscientiousness, were able to predict both task performance and OCB to varying

degrees of success (Borman & Motowidlo, 1993).

It is not surprising that job satisfaction and personality were able to predict

OCB, to a certain degree. The job performance model originally developed by Campbell

(1990) and then extended by Campbell et al. (1993) provided some understanding as to

why these two variables were pivotal OCB antecedents. Consider the three performance

components when the performance was OCB. Generally, declarative knowledge

consists of facts, principles, goals, and self-knowledge; while procedural and knowledge


and skills includes cognitive skills, psychomotor skills, physical skills, self-management

skills, and interpersonal skills. When considering these two performance components

from an OCB perspective, the individual would need to know the behaviours that

constituted OCB and then how to apply that knowledge and behaviour in the workplace.

Campbell (1990) recognised that both of these performance components could be

predicted by aspects of ability, personality, interests, education, training, experience,

and aptitudes/treatment interactions. Additionally, motivation was the third performance

component and it considers the effort of the individual to perform OCB. The main role

of job satisfaction in the production of OCB would involve attitudinal aspects that

would increase the individual’s involvement, which was related to the motivation

component. In contrast, as a disposition, personality would take a more complex role

across all three determinants. Specifically, the personality factors of conscientiousness

and extraversion have been the most consistent OCB antecedents, while agreeableness

and neuroticism have also been observed to have small roles in the prediction of OCB.

Moreover, conscientiousness has also been a consistent predictor of task performance.

Organ et al. (2006) noted that past studies have argued that job satisfaction had

some bearing on the production of OCB based on attitudinal factors, while personality

influenced the production of OCB based on dispositional aspects. Consistent with this

argument, Organ and McFall (2004) suggested that personality predictors were useful in

the selection and placement processes within the organisation, while attitudinal

antecedents, such as job satisfaction, were important for managers to ensure that their

staff remained satisfied in order to produce OCB. However, while both these variables

were considered pivotal predictors of OCB, the correlations have been small to

moderate. Additionally, their roles may be indirect and mediated through other

variables, such as job involvement (e.g. Diefendorff et al., 2002; and Rotenberry &


Moberg, 2007). Subsequently, additional variables have been considered in an attempt

to gain a more thorough understanding of OCB prediction. The main thrust of this

research attention has been directed at attitudinally-based variables, such as affective

disposition. Additionally, as declarative knowledge was involved as a performance

determinant of OCB, researchers have returned their attention to the possibility of

general mental ability having a role in the prediction of OCB through declarative

knowledge and procedural knowledge and skills that were required to produce OCB.

One such skill highlighted in Campbell’s et al. (1993) model was interpersonal skill.

Interpersonal skills exist within the procedural knowledge and skills performance

component and are argued to be highly relevant for OCB. Notably, Campbell et al.

suggest this behaviour is also required across the eight common performance

components. The additional variables of affective disposition and interpersonal skills

provide the opportunity to investigate a more complex understanding of the prediction

of OCB, through the study of multiple antecedents.

Affective Disposition

Affective disposition has been a variable that was investigated in the OCB

domain and was generally studied in conjunction with job satisfaction and/or

personality. Affective disposition theory suggests that individuals possess an affective

disposition which can affect their mood (Watson, Clark, & Tellegen, 1988). When this

theory is applied to the work environment, an employee’s working behaviour and

subsequently the likelihood of performing OCB, can be influenced by their affective

disposition via the mood of the employee.

Watson et al. (1988) argued that the affective structure consisted of two

orthogonally related factors, existing in both trait and situational forms. These two


broad dimensions were positive affect and negative affect. Positive affect reflected a

high level of energy that involved the feelings of enthusiasm, activity, alertness, high

concentration, and pleasurable engagement. In contrast, individuals with a low level of

positive affect were characterised by sadness and lethargy (Watson et al., 1988).

Alternatively, Watson et al. suggested that negative affect represented adverse mood

states including anger, contempt, disgust, guilt, fear, and nervousness. Moreover, these

mood states existed within a general dimension of subjective distress and unpleasurable

engagement. Alternatively, individuals with low negative affect were considered to be

in a state of calmness and serenity (Watson et al., 1988).

Similar to personality, Watson and Clark (1997) stated that the affective

disposition structure was hierarchical with the two broad dimensions of positive and

negative affect consisting of specific facets. Notably, they suggested that research to

identify the broader more general dimensions of affective disposition has observed

stronger evidence than the more specific facets. In the conclusion of their literature

review, Watson and Clark suggested that a stable taxonomy has yet to be identified for

the lower order, specific affect level. However, as noted earlier by Watson and Clark

(1992a), both levels of the hierarchical structure were important for the comprehensive

measurement of affective disposition.

One of the most common scales to measure affective disposition is the Positive

and Negative Affect Scale (PANAS). This scale was developed by Watson et al. (1988)

in an attempt to improve the measurement reliability and validity of affective

disposition. The scale consisted of 20 items and measured positive and negative affect,

with 10 items per dimension. Watson et al. (1988) argued that the scale possessed high

internal consistency and demonstrated stability over a two month time period.


Importantly, Watson and Clark (1997) noted that the PANAS has reliably measured the

two broad dimensions of affective disposition across a wide variety of conditions.

Affective Disposition and Personality

Personality contains different aspects of trait versus state throughout its structure

(see Eysenck, 1998; Maddi, 2001). For example, the five factors of personality and

affective disposition both form part of the individual’s personality yet exist within

different areas of this structure. The five factors of personality are broad dimensions and

are contained in the core of personality. These dimensions were inherited and only

changed due to maturation (Maddi, 2001). In contrast, affective disposition could occur

at the trait or state level of personality, which could involve aspects of heritability while

other elements were learned through environmental influences, such as culture and

family. In its state form, affective disposition could be influenced by specific situations

and circumstances (George, 1991; Watson et al., 1988).

Affective disposition had been observed to relate to both broad personality

dimensions and also their lower order facets. In an early study, Costa and McCrae

(1980) investigated the relationship between personality and the construct of happiness,

alternatively known as subjective well-being. The authors obtained happiness ratings

using the Bradburn Scales, which consisted of two affective disposition scales that

measured positive and negative affect. The two scales were summed to produce a score

for an overall measure of happiness. The scales were mailed to 1100 males four times at

three monthly intervals in 1976. Response rates for each mailing were 79%, 82%, 73%,

and 54%, respectively. In addition to affective disposition, Costa and McCrae (1980)

collected data for hopelessness, personal security, life satisfaction, temperament, and the

personality factors of extraversion and neuroticism. The authors proposed that as


positive and negative affect were considered independent dimensions that contributed to

an overall measure of happiness, the personality antecedents for each dimension would

be different. The authors proposed a model to explain the relationships (Figure 3.4).

Results suggested that positive affect and negative affect were directly related to

extraversion and neuroticism, respectively. They also proposed that extraversion and

neuroticism had an indirect effect on global happiness, which was related to morale, life

satisfaction, hopelessness, and affect balance. It is noted that the study did not employ

causal analysis and consequently, the model requires additional testing. However, the

study provided a useful basis for OCB antecedent research, as both job satisfaction and

morale, as discussed by Organ and Ryan (1995), were considered important in the

prediction of OCB.

Figure 3.4

Personality influences on subjective well-being via affect (Costa & McCrae, 1980, p.675).

The relationships between affective disposition and both extraversion and

neuroticism have been explained by Eysenck and Eysenck (1985). Firstly, it is

important to consider the biologically-based differences as proposed by Eysenck (1967).

Eysenck argued that the two personality factors of extraversion and neuroticism were

orthogonal personality dimensions that corresponded to different biological bases. In his

model, the extraversion dimension was characterised with different levels of activity in

Neuroticism:

Anxiety

Hostility

Impulsmty

Psychosomatic Complaints

Negative Affect

Dissatisfaction

Extraversion:

Sociability

Tempo

Vigor

Social Involvement

Positive Affect:

Satisfaction Subjective Well-Being

“Happiness”:

Morale

Life Satisfaction

Hopefulness

Affect Balance


the corticoreticular loop. Extraverts would have lower levels of cortical activity as

compared to introverts. In contrast, Eysenck and Eysenck suggested that the neuroticism

dimension differed from variations in the limbic system, which controlled emotion.

Individuals high in neuroticism would have a more active limbic system than those low

in neuroticism. As such, Eysenck and Eysenck suggested that extraverts would seek

additional stimulation through social interaction and excitement. This would lead to

positive emotions and then extraversion would be related to positive affect. In contrast,

individuals high in neuroticism would have a high level of emotional response and be

more likely to be related to negative affect.

Rusting and Larson (1997) observed support for Eysenck and Eysenck’s (1985)

model of the relationships between the dimensions of affective disposition and the

personality traits of extraversion and neuroticism. Their study assessed the mood state

of 150 participants who read written scenarios that were designed to induce either

positive or negative affect. Measures of personality and initial mood state were taken

prior to the imagination task. Following the imagination tasks, mood was measured.

Participants completed both positive and negative tasks that were separated by a neutral

mood activity. Results indicated that extraversion was positively related to positive

mood when pleasant imagery occurred. In contrast, neuroticism was positively related

to negative mood when the unpleasant imagery task was shown. The relationships

between extraversion-positive affect and neuroticism-negative affect were also observed

when mood was controlled. The results by Rusting and Larson were consistent with the

model developed by Eysenck and Eysenck’s (1985). Specifically, Rusting and Larson

suggested that positive affect and negative affect were two separate affect dimensions

that were related to extraversion and neuroticism, respectively.


The high correlations between affective disposition, extraversion, and

neuroticism have prompted some researchers to question if positive affect and negative

affect were equivalent constructs to extraversion and neuroticism, respectively. In

contrast, Nemanick and Munz (1997) suggested while they were dispositional traits,

personality and affective disposition were hierarchically structured. Specifically, as

previously mentioned, Eysenck and Eysenck (1985) argued that personality traits could

influence the individual’s mood state across many different situations. This position was

also consistent with the personality structure as described by Eysenck (1998) and Maddi

(2001). In other words the personality traits of extraversion and neuroticism were

considered traits that represented broad and general characteristics, while affective

disposition was a narrower trait and closer to a state level. Moreover, these authors

contend that affective disposition was more able to impact behaviour at any point in

time. To assess their theory, Nemanick and Munz (1997) conducted a repeated measure

investigation utilising the responses of 103 psychology students using survey measures

of extraversion, neuroticism, and trait mood. Results supported a mediation relationship

between trait mood and the personality traits of extraversion and neuroticism.

Specifically, when utilising the PANAS as the measure of affective disposition, positive

affectivity demonstrated a mediator role between extraversion and state positive affect,

while negative affectivity demonstrated a mediator role between neuroticism and state

negative affect. Personality and affective disposition could thus be seen as related but

sufficiently different constructs, as was argued by Eysenck and Eysenck (1985).

In the study of personality and emotion, the extraversion-positive affect

relationship has been described as an extremely robust finding (Lucas, Le, &

Dyrenforth, 2008). While the relationship has been observed consistently in research

(e.g. Costa & McCrae, 1980; Lucas et al., 2008; Nemanick & Munz, 1997; and Watson,


Clark, McIntyre, & Hamaker, 1992), there is some debate if the relationship was direct

or mediated by additional processes. To investigate this issue, the authors conducted a

study to determine if extraversion was temperate and therefore had a direct relationship

with positive affect; or conversely, if social activity mediated the relationship between

extraversion and positive affect; and/or extraversion mediated the relationship between

positive affect and the social situation. Using two studies to replicate the results, the

first study had a population of 144 and utilised self-report measures for extraversion and

positive affect and three weeks of diary-based assessments in addition to seeking

external reporters providing data for the participant’s positive affect. The social

situation data gained for the first study was based on general social activity and one

specific measure of going to parties. Data were analysed using regression analysis,

structural equation modeling, and hierarchical linear modeling techniques. The

structural equation modeling demonstrated a significant moderate causal pathway from

extraversion to positive affect of 0.45. Extraversion also held small but significant

causal pathways to all of the social activity variables (Friend: 0.32; Leading: 0.27;

Helping: 0.20; Alone: -0.24). Further analysis demonstrated that extraverts engaged in

more social activities than introverts. When the amount of social activity was controlled

for, the results suggested that the association between extraversion and positive affect

was direct. The second study changed the measurement of social situation to specific

social activities. Lucas et al. (2008) gained data from a further 186 participants and the

results supported those obtained in the first study. Furthermore, the second data set

indicated that there was no difference between the positive reaction by extraverts and

introverts when controlling for specific interaction partners and types of social

activities.


Research to investigate possible relationships between an individual’s affective

disposition and the related personality traits have not been confined to extraversion and

neuroticism. To consider the relationships between the trait levels of personality and

affective disposition, in addition to their more specific facets, Watson and Clarke

(1992b) measured affective disposition and personality factors across four sample

groups (n1=532, n2=236, n3=225, n4=325). The authors utilised the PANAS and various

measures of personality (including scales developed by Goldberg, 1983, NEO-FFI, and

NEO-PI) to measure the broad trait dimensions of both constructs. The NEO-PI also

measured the specific facets of the broader personality dimensions. Specific facets of

positive and negative affect were measured using the expanded form of the PANAS,

developed by Watson and Clark (1990; as cited in Watson & Clark, 1992b). Seven of

the strongest and most consistent markers of positive affect (joviality, self-assurance,

and attentiveness) and negative affect (fear, sadness, guilt, and hostility) were

considered. The results of their study reinforced the strong and pervasive associations

between extraversion and positive affect (r=0.58), and neuroticism and negative affect

(r=0.58). Notably these correlations occurred at both the broad levels, and also the

specific facet levels. Small to moderate correlations were also observed for positive

affect and conscientiousness (r=0.37), openness (r=0.33), and agreeableness (r=0.21).

Watson and Clark (1992b) suggested that some of the weaker correlations

between the personality factors and affective disposition may be due to the five factors

not being completely independent of each other. As such, they conducted multiple

regression analysis using the five personality factors as predictors against the two

criteria of positive and negative affect. Results supported the relationships between

negative affect and neuroticism and positive affect and extraversion. Specifically, for

negative affect, only neuroticism contributed meaningfully to the variance across the


samples (27.5% - 41.8%). A more complex pattern was observed for positive affect, as

both extraversion (24.3% - 58.2%) and conscientiousness (5.5% - 22.0%) contributed to

the variance of positive affect. Watson and Clark stated that extraversion had a stronger

relationship across the specific markers of positive affect than conscientiousness, with

the exception of attentiveness. The authors suggested that extraversion was broadly

related to positive affect, while conscientiousness was specifically related to positive

affect through attentiveness. Furthermore, when the facets of conscientiousness were

assessed with the broader dimension of positive affect, the strongest correlation was

observed for achievement (r=0.44), while small correlations were also observed with

dependability (r=0.23) and orderliness (r=0.13). As such, while extraversion had

consistently demonstrated a strong relationship with positive affect at broad and facet

level, these results suggested that conscientiousness also had a relationship with positive

affect, albeit smaller.

In addition to the strong relationship that was observed between both broad and

specific facets of negative affect and neuroticism, Watson and Clark (1992b) observed a

relationship between negative affect and agreeableness (r=-0.23). Following multiple

regression, the authors explained this relationship through the negative affect facet of

hostility possessing a negative relationship with agreeableness (r=-0.33). They noted

that the measure of hostility contained a strong, interpersonal component, which they

suggested contributed to the relationship between hostility and agreeableness.

The study conducted by Watson and Clark (1992b) was successful in

demonstrating relationships between the broad affective dispositions and personality

dimensions, particularly for positive and negative affect with extraversion and

neuroticism, respectively. Moreover, this study also demonstrated evidence for the

importance of facets within the conscientiousness and agreeableness factors to possess


relationships with lower order affective disposition facets. As such, the study reinforced

that the strongest personality relationship for positive affect was extraversion with a

smaller correlation for conscientiousness; while negative affect had a stronger

relationship with neuroticism and also a small correlation with agreeableness when

negative affect was measured using interpersonal aspects.

While affective disposition and personality were acknowledged as related

constructs, researchers considered that their impact on performance, and more

specifically OCB, was indirect at a trait level (e.g. George, 1991; George & Brief, 1992;

Organ & Ryan, 1995). Evidence for this assumption was provided in a study by George

(1991) who investigated the relationship between state and trait mood with discretionary

and non-discretionary behaviour. Her results from 221 salespeople indicated that

positive mood increased the occurrence of helping behaviours that were discretionary or

prescribed. State positive mood demonstrated small positive correlations with altruism,

a type of OCB (r=0.24) and customer service, a form of task performance (r=0.26).

Conversely, a significant correlation was not observed between trait mood and job

performance. The author maintained that positive mood and more specifically positive

affect, influenced the cognitions and behaviour of the employee. Positive affect may

impact the positive mood state, while state mood was also influenced by other factors,

such as situation and personality.

Affective Disposition, Job Satisfaction, Personality, and Organisational Citizenship

Behaviour

The indirect role of affective disposition in the prediction of OCB resulted in

researchers studying affective disposition in conjunction with job satisfaction or broader

factors of personality, rather than directly predicting OCB. Investigating OCB through a


combination of affective disposition and job satisfaction was a reasonable approach

given that job satisfaction related to an emotional state. Consequently, affective

disposition could be a dispositional determinant of job satisfaction, as contended by

George (1992). Moreover, affective disposition could be studied in both state and trait

forms, with the trait form holding the most interest for personality researchers.

Hochwarter, Perrewé, Ferris, and Brymer (1999) found that affective disposition

was an important variable in understanding the job satisfaction-performance

relationship within a motivational context. They surveyed 270 participants who held

managerial jobs in national hotel chains across the United States of America. Self-report

measures were utilised for job performance, job satisfaction, affective disposition, and

value attainment – a motivational construct. The performance measure consisted of both

task behaviours and OCB. In addition to the performance self-report measure,

participants were requested to provide their most recent performance appraisal score.

The two performance measures were combined to produce a single score for

performance. Results indicated that performance was related to job satisfaction

(r=0.22), and positive affect (r=0.35) but not negative affect. This result suggested that

positive affect had a stronger role in the prediction of performance than negative affect.

Additionally, job satisfaction was related to both positive affect (r=0.30) and negative

affect (r=-0.36). Hierarchical regression analysis showed a three-way interaction

between job satisfaction, value attainment, and affective disposition in the prediction of

performance. Specifically, the interaction that demonstrated the strong positive

relationship between job satisfaction and performance occurred when high value

attainment was present with either high positive affect or low negative affect. As such, a

positive mood combined with strong motivation strengthened the job satisfaction-

performance relationship.


The role of affective disposition in the job satisfaction-performance relationship

continued to interest researchers. Recently, Kaplan, Bradley, Luchman, and Haynes

(2009) conducted a meta-analytic study investigating the role of trait positive and

negative affect in job performance for both OCB and task performance. Other aspects

studied included withdrawal behaviour (absenteeism, lateness, turnover, and tardiness),

counterproductive work behaviour (aggression, violence, bullying, theft, sabotage, and

harassment), and occupational injury. Incorporating 57 studies, their results indicated

that positive affect was correlated with OCB (r=0.23) but not to withdrawal behaviours;

while negative affect had a small negative correlation with OCB (r=-0.10) and

positively correlated with withdrawal behaviours (r=0.16), counterproductive work

behaviours (r=0.30) and occupational injury (r=0.20). Positive affect and negative affect

were observed to have small correlations with task performance (rPA=0.19; rNA=-0.15),

stronger relationships with self-rated performance (rPA=0.35; rNA=-0.30), while no

relationship was observed for the objectively rated performance. This result may have

indicated inflation of the correlation due to common method bias. However, consistent

with previous research, mediation analysis indicated that affect influenced a variety of

behaviours through different mechanisms. Consequently, the authors argued that

dispositional traits, such as positive affect, did influence job behaviour, both in respect

of OCB and task performance.

While the studies of Hochwarter et al. (1999) and Kaplan et al. (2009)

demonstrated the importance of job satisfaction and positive affect in the prediction of

OCB, personality variables were not considered by either team of researchers. Notably,

Hochwarter et al. (1999) argued that although their study was the first logical step in the

analysis of how positive affect and value attainment impacted the job satisfaction-

performance relationship, further research simultaneously studying multiple variables


was needed to consider the complexity of job performance. Moreover, these authors

argued that research examining the relationship between personality and performance

was particularly relevant given that positive and negative affect have also been shown to

relate to specific dimensions of the five factor model of personality.

In their meta-analytic study Organ and Ryan (1995) developed a model to test

hypothesised relationships between multiple antecedent variables. These variables

included: job satisfaction, demographic, personality, knowledge, skills, ability,

incentives, and contractual rewards to predict OCB and task performance. Their model

is reproduced in Figure 3.5 and displayed two solid lines from the hypothesised

antecedents of OCB and task performance. The solid causal pathways represented the

strongest relationships. However, there were some situations in which an OCB

antecedent could also be an antecedent of task performance. For example, the

personality trait of conscientiousness had been observed to predict performance across a

wide variety of positions; and extraversion has also predicted performance when the job

had interpersonal aspects, such as sales performance (e.g. Barrick & Mount, 1991). The

second type of causal relationship was considered a minor pathway and Organ and Ryan

represented this weaker pathway by dotted directional lines.


Organ and Ryan (1995) suggested that there were potential moderators to the

job attitudes/disposition/personality-OCB relationship, with the strongest being

identified as the use of self-report versus other-ratings of OCB (Figure 3.5). While the

authors proposed self-report measures would produce inflation of the mean correlation,

they also suggested a stronger concern was the possible instability of the response due

to temporal or situational moderators. For example, if a satisfaction item required a

response regarding a recent pay reduction, the individual may feel that they have to

justify their dissatisfaction by overstating the response. The temporal or situational

moderators were considered a stronger moderator than other aspects including age,

tenure, rank, or type of work.

JOB ATTITUDES

DISPOSITION/PERSONALITY

KNOWLEDGE

SKILLS

ABILITY

x

INCENTIVES

CONTRACTUAL REWARDS

ORGANIZATIONAL

CITIZENSHIP

BEHAVIOR

TASK

PERFORMANCE

POTENTIAL

MODERATOR

S

Figure 3.5

Hypothesised correlates of organisational citizenship behaviour versus in-role task

performance, as reproduced in Organ and Ryan (1995, p. 777).


The previous discussion has suggested that OCB antecedents were attitudinal or

dispositional in nature. The most consistently demonstrated relationship was between

extraversion-positive affect (e.g. Costa & McCrae, 1980). Notably, while the

neuroticism-negative affect relationship was also considered consistently strong, the

evidence was not sufficiently strong enough to suggest negative affect held a significant

role in the prediction of OCB (e.g. Hochwarter et al., 1999). In contrast,

conscientiousness has demonstrated relationships with positive affect, job satisfaction,

job involvement, OCB, and overall performance (e.g. Barrick & Mount, 1991;

Diefendorff et al., 2002; Organ & Ryan, 1995; Watson & Clark, 1992b). While results

for the antecedents-OCB relationship have been small to moderate, there has been

sufficient evidence gained to propose that a range of variables have a role in the

prediction of OCB. The multiple relationships that have been shown to predict OCB are

summarised at Figure 3.6.

Job

Satisfaction

Extraversion

Conscientiousness

Positive

Affect

Job Involvement OCB

OCB antecedent researchers have focused initially on dispositions and

attitudinal variables that would contribute to the motivational pathway, as discussed by

Campbell (1990) and Campbell et al, (1993). However, more recently there has been a

Figure 3.6

Summary of hypothesised pathways for the motivationally-based variables of personality,

positive affect, and job satisfaction.


renewed interest to investigate what specific abilities and aptitudes may have a role in

the prediction of OCB. Campbell’s performance determinants of declarative knowledge

and procedural knowledge and skills provided a useful framework for this investigation,

as suggested by Brown and Leigh (1996). Investigating ability-related variables in

addition to motivation-related variables could enhance our knowledge and

understanding of OCB antecedent research by evaluating multiple predictors

simultaneously. The next section of this chapter discusses the variable of ability and its

role in OCB prediction.

Ability

Ability testing has been a central focus of research for organisational

psychologists for over 100 years. Charles Spearman is recognised for defining the

construct of general mental ability in 1904 (Schmidt & Hunter, 2004). Other major

researchers and theorists in the field of ability testing are Galton (1884) – investigated

sensory and perceptual skills; Binet and Simon (1905) – development of the Binet-

Simon intelligence test; and Terman (1916) – development of the Standford-Binet

intelligence test (as cited in Grubb, Whetzel, & McDaniel, 2004; Doverspike et al.,

2004). General mental ability testing was shown to be the best procedure for predicting

job related learning, specifically ability to acquire knowledge on the job and

performance in training programmes (Grubb et al., 2004). Most measures of general

mental ability assessed aspects of numerical, verbal, and/or spatial ability.

In a review of almost 100 years of general mental ability research, Schmidt and

Hunter (2004) argued that general mental ability predicted job performance better than

other abilities, traits or dispositions. These authors contended that while other traits,

such as personality, were recognised to influence behaviour, the correlations between


these traits and job performances were not as strong as the general mental ability-

performance relationship. Furthermore, Schmidt and Hunter (2004) stated that the

correlations between general mental ability and later occupational level, training

performance, and job performance are in excess of 0.50, which was rare and considered

strong in psychological research. These results have been reproduced across a range of

occupations (Gatewood & Field, 1998). Given these findings, general mental ability has

remained critical in the prediction of job performance.

Importantly, despite Schmidt and Hunter’s (2004) strong conviction that general

mental ability was the best predictor of performance, they also acknowledged that other

traits, such as personality, when researched simultaneously with general mental ability,

could increase the overall validity of the prediction. Moreover, the multidimensional

nature of performance had enabled researchers to theorise that different types of

performance would have different antecedents. As discussed previously, Motowidlo and

Van Scotter (1994) observed that task performance was more related to experience and

ability, while OCB had stronger relationships with individual traits, such as personality.

The distinction between antecedents provided evidence of the multidimensional nature

of performance.

Despite being useful for investigating the multidimensionality of performance

through the study of different antecedents, the distinction of task versus OCB has

restrained some researchers from investigating the specific abilities that could be used to

predict OCB. However, this trend has now changed. While it is acknowledged that the

relationships are stronger for task performance-ability and OCB-motivational aspects,

researchers have also began to consider that OCB may have some ability-related

antecedents. The existence of OCB having both ability and motivationally based

antecedents was hypothesised by Organ and Ryan (1995), as summarised earlier in


Figure 3.5. Motowidlo, Borman and Schmit (1997) also theorised similar relationships,

which are summarised at Figure 3.7. Note that while Motowidlo et al. discussed

contextual performance in their article, this construct has been referred to as OCB for

the purpose of the current study.

In the model at Figure 3.7, Motowidlo et al (1997) suggests that OCB and task

performance would be influenced through various characteristic adaptations. These

adaptations were considered to be the basic tendencies that were learnt through

experience during the interaction of ability and personality across situations. Personality

variables, such as extraversion, conscientiousness, and agreeableness would influence

Personality

Variables

Cognitive

Ability

Variables

OCB

Habits

OCB

Skill

OCB

Knowledge

Task

Habits

Task

Skill

Task

Knowledge

OCB Task

Performance

Figure 3.7.

Motowidlo et al. (1997) theory of individual differences in task performance and OCB.


OCB through learnt OCB habits, skills, and knowledge, while also influencing task

habits, albeit to a lesser degree. Conversely, cognitive abilities would influence task

performance through task habits, skills, and knowledge. To a lesser extent, cognitive

abilities were also theorised to influence OCB through OCB knowledge and

understanding when to apply skills.

Motowidlo et al.’s (1997) model is consistent with the performance model that

was developed by Campbell (1990) and extended by Campbell et al. (1993).

Specifically, Motowidlo et al. argued that knowledge consisted of facts, principles, and

procedures that when applied to an OCB performance, would include situations that

involved helping, cooperating, endorsing, supporting, defending, persisting in effort,

and volunteering. Moreover, OCB skills would involve the practical application of the

knowledge in these situations. As such, personality traits that assisted interpersonal

relationship, such as agreeableness and extraversion, would be expected to influence

OCB. Additionally, conscientiousness would be theorised to improve OCB through

increased levels of effort across challenging OCB situations, including interpersonal

relationships. Furthermore, Motowidlo et al. considered that cognitive ability would

enhance the individual’s acquisition and practical application of the OCB knowledge

and skills. As such cognitive ability, albeit to a less degree would influence the

production of OCB. The third characteristic adaptation involved work habits that were

influenced by both individual differences through motivational aspects and also learned

behaviour. From an OCB perspective, these habits included characteristics of how the

individual handled conflict, what was the preferred communication style or

interpersonal interaction style. The habits are suggested to consist of situationally-

relevant behaviour, which could be influenced by the broader traits of personality.


Notably, throughout all the theoretical models that have been discussed in this

thesis (e.g. Campbell, 1990; Campbell et al., 1993; Organ & Ryan, 1995; and

Motowidlo et al., 1997), interpersonal skills have been highlighted as the most

important ability construct for OCB. As discussed, these models suggested that

cognitive ability, such as general mental ability had an indirect role in the production of

OCB through skills such as interpersonal skills. Importantly, this skill was identified as

one of the procedural knowledge and skills by Campbell et al. (1993), in addition to

existing within many of the eight major performance components for performance that

required interpersonal interactions, as previously discussed in this chapter. As such, the

following section considers the role of interpersonal skills in the prediction of OCB.

Moreover, as will be discussed, general mental ability has been shown to be related to

interpersonal skills, as theorised previously, in addition to personality traits.

Interpersonal Skills

Almost all jobs require individuals to interact with other personnel to complete

their required tasks. Individuals across positions and organisational levels may have to

deal with people including but not limited to: clients, supervisors, peers, subordinates,

and support staff. Researchers have been increasingly considering the importance of

interpersonal skills as a predictor in job performance (Ferris, Witt, & Hochwarter,

2001). Based on the work of Argyle (1969), Meichebbaum, Butler, and Gruson (1981),

and Gardner (1993), Ferris et al. (2001) defined interpersonal skills as the “interpersonal

perceptiveness and the capacity to adjust one’s behavior to different situational demands

and to effectively influence and control the responses of others” (p. 1076). Major

interpersonal skills areas included: communication – listening and responding to verbal

and non-verbal communication, questioning, and presenting information; motivation –

goal setting, coaching, counseling, mentoring, and delegation; leading – helping,


facilitating, politicking, persuading, managing change; teaming – working in teams,

conducting meetings, and valuing diversity; and problem solving – ethical decision

making, conflict resolution, and negotiation (Hayes, 2002; Robbins & Hunsaker, 2006).

Interpersonal skills and general mental ability are considered important

attributes of employees in the work environment. Ferris et al. (2001) demonstrated that

both variables were required to predict high levels of performance when they

investigated their interactive effects. Importantly, their study utilised a performance

scale that measured job dedication and interpersonal facilitation, thereby measuring

aspects of OCB. In addition to OCB performance, measures of personality, social skills,

general mental ability, salary and demographics were utilised. Performance was

measured using supervisor reports, while other variables were measured using self-

report surveys. The authors collected complete data on 106 programmers in a systems

development organisation. Results indicated that personality and social skills were

related (r=0.20), a finding that has also been observed by Hogan (1991). Moreover, they

suggested that individuals who possessed a high score for their general mental ability

and also a high level of social skills produced more OCB. While general mental ability

had traditionally been related to task performance, given these results, it was argued that

general mental ability may also be a predictor of OCB through specific skills, such as

interpersonal or social skills. Brown (1996) argued that cognitive ability would equate

to working smart (declarative knowledge and procedural knowledge and skills) as

compared to working hard (motivation). This argument was further supported when

applied to Campbell’s performance model (1990), as all three determinants were

required to produce any type of performance. While the discretionary nature of OCB

would be likely to require a higher level of motivation than non-discretionary

behaviour, such as task performance, OCB specific declarative knowledge and


procedural knowledge would also be required by the individual to produce OCB.

However, as suggested by Organ and Ryan (1995), the relationship between general

mental ability and OCB would be weaker than the general mental ability-task

performance relationship.

While the relationship between social skills and general mental ability had

produced some promising results, researchers have also investigated the personality-

social skills relationship. As previously mentioned, Mount et al. (1998a) analysed 11

studies (N=1586) in a meta-analytic study investigating the five factor model-

performance relationship for jobs involving interpersonal interactions. The 11 studies

used the unpublished Personal Characteristics Inventory that was developed in 1995 by

Mount and Barrick to measure personality. Performance was assessed using supervisor

rating on overall performance and interpersonal interactions. Analysis was conducted

between the five factors of personality and performance in team-work (T) versus dyadic

(D) positions. The authors demonstrated that emotional stability and agreeableness were

more strongly related to team-work related employment than to positions that directly

provided a service to customers or other employees for both types of positions

(emotional stability: overall performance – rT=0.27, rD=0.12, interpersonal interactions

– rT=0.25, rD=0.14; agreeableness: overall performance – rT=0.33, rD=0.13,

interpersonal interactions – rT=0.35, rD=0.22). Their results also indicated that

conscientiousness was positively related to both types of jobs for: overall performance

(rT=0.21; rD=0.29) and interaction performance (rT=0.17; rD=0.23). Extraversion and

openness to experience were related to a lesser degree with the strongest relationship

occurring when extraversion was compared to overall performance in a team

environment (rT=0.22, rD=0.07). These results suggested that personality was related to

social skills in a workplace setting.


While the results for personality-interpersonal skills relationship have been

receiving research attention in the prediction of performance, not all studies have

demonstrated the relationship consistently across all five factors. Notably, as mentioned

previously, Barrick and Mount (1991) demonstrated that while conscientiousness

predicted performance across all types of job performance and extraversion predicted

job performance across occupations that involved social interaction, agreeableness did

not demonstrate a relationship with job performance. The insufficient evidence to

support the agreeableness-interpersonal skills relationship was also observed by Barrick

et al. (2002), as discussed previously.

These results would suggest that no relationship existed between agreeableness

and interpersonal skills. Nevertheless, the existence of this relationship has been

demonstrated, especially where the performance is team related. The inconsistent results

are similar to many of the relationships in performance prediction studies. In one such

study, Halfhill, Neilsen, and Sunstrom (2008) investigated the relationship between

group performance in military action teams and the personality factors of

conscientiousness (attention to detail, dependability, and reliability) and agreeableness

(avoid conflict, group cohesion, and interpersonal). The authors posited that

agreeableness was an important trait that was related to group effectiveness, specifically

due to the interpersonal nature of group work. Alternatively, Halfhill et al. contended

that conscientiousness was important due to the task orientation of the individual. The

final data sample consisted of 166 cases and correlational analysis was conducted

between the mean and minimum levels of agreeableness and conscientiousness. Results

indicated that the mean and minimum levels of agreeableness correlated positively with

group performance evaluations (r=0.32; r=0.40, respectively), while minimum levels of

conscientiousness produced a positive correlation with group performance (r=0.39). The


study by Halfhill et al. (2008) supported the results observed by Neuman, Wagner, and

Christiansen (1999) and also Neuman and Wright (1999), who conducted similar

studies using populations from the retail and human resource teams, respectively. On

the basis of these three studies across action teams from different job groups, Halfhill

and his colleagues suggested that agreeableness was related to the successful

collaboration in the groups.

The existence of a relationship between agreeableness and interpersonal skills

was also supported by Witt, Burke, Barrick, and Mount (2002). The authors

investigated the possibility of an interactive effect between conscientiousness and

agreeableness in explaining job performance. Specifically, they hypothesised that a

stronger relationship would exist between conscientiousness and job performance when

the individual had a higher level of agreeableness. Data were collected from one public-

sector and six private-sector organisations. Notably, the correlation between

agreeableness and supervisors rating for the five samples that required co-operative

interaction as part of the job was 0.15, while the correlation for the two samples that did

not require co-operative interaction was non-significant. Conscientiousness was

correlated with job performance across all sample groups ranging between 0.16 and

0.28. Using the sample groups that required cooperative interaction further analysis was

conducted to assess the interaction between conscientiousness and agreeableness. The

results from Witt and colleagues were consistent with their hypothesis. Namely the

conscientiousness-performance relationship was stronger when agreeableness was high,

as compared to when agreeableness was low. Moreover, if the worker had a low score

for conscientiousness, the agreeableness score was unrelated to the performance rating.

The authors concluded that agreeableness was required for conscientious employees to

perform effectively when interpersonal interactions were part of the job. In contrast, if


the employee was not conscientious, the level of agreeableness was not relevant for the

prediction of performance.

In considering the results of the previous discussion, the main three personality

factors that are of interest to interpersonal skills researchers are conscientiousness,

extraversion and to a lesser degree, agreeableness. As stated earlier in this chapter, John,

(1990) described conscientious employees as organised, thorough, planful, effective,

and responsible, while individuals with high scores on extraversion were considered

talkative, assertive, active, energetic, outgoing, outspoken, and dominant. High scores

on the agreeableness factor were described as sympathetic, kind, appreciative,

affectionate, soft-hearted, warm, and generous (John, 1990). These sets of personality

descriptions would be highly relevant to predict OCB for individuals possessing high

levels of competence for interpersonal skills, in conjunction with general mental ability.

A summary to predict OCB through interpersonal skills and its main antecedents is

proposed in diagrammatic form at Figure 3.8.

Extraversion

Conscientiousness


General Mental Ability

Agreeableness

OCB

Figure 3.8

Summary of causal pathways to predict OCB through ability (interpersonal skills) using the

variables of general mental ability and personality.


Conceptual Model to Predict Organisational Citizenship Behaviour

The prediction of OCB continues to interest Industrial-Organisational

researchers. OCB was studied initially using job satisfaction as the main predictor,

while personality has also been increasingly investigated as an antecedent of OCB.

These two constructs are theorised to be antecedents of OCB as they influence the

motivation of the employee to produce non-discretionary behaviour in the workplace.

More recently, researchers have begun to investigate other variables in addition to the

two pivotal antecedents. All the major researched variables and their relationships will

now be incorporated into a single model in order to investigate the main antecedents of

OCB simultaneously.

Building the Model

The conceptual framework is captured in the title of the thesis, namely:

Predicting organisational citizenship behaviour – More than job satisfaction and

personality. While job satisfaction and personality have received the majority of the

early research focus into the prediction of OCB, as has been discussed, results have

remained inconsistent. The simple relationships of either job satisfaction and OCB or

personality and OCB have been inadequate to capture the complexity of OCB in the

work environment. Consequently, additional investigations have been conducted to

identify other important variables that would assist in the prediction of OCB.

Positive affect has been identified as an important attitudinal variable that may

have an indirect role in the prediction of OCB (see George, 1991, 1992; George &

Brief, 1992; Hochwarter et al., 1999; and Kaplan et al., 2009). Furthermore, OCB

prediction modeling has focused on motivational aspects while largely ignoring specific


abilities. Notably, this trend has been changing, as OCB researchers increasingly

recognise that there may be abilities that would promote OCB in the workplace.

Specifically, researchers have begun to re-use the more traditional performance-related

variable of general mental ability and specific OCB related abilities, including

interpersonal skills. Campbell’s et al. (1993) performance model identifies interpersonal

skills as an important skill when the job component involves aspects of interpersonal

interaction. As such, interpersonal skills are relevant for the production of OCB, which

is considered to consist of interpersonal interactions (Organ et al, 2006). To identify the

specific abilities and dispositional traits that would be important antecedents of

interpersonal skills, researchers have considered personality traits and general mental

ability. General mental ability is one of the best predictors of overall job performance

(e.g. Doverspike et al., 2004; Gatewood & Field, 1998; and Schmidt & Hunter, 2004)

and also interpersonal skills (e.g. Ferris et al., 2001). Moreover, Ferris et al. (2001)

observed a relationship between interpersonal skills and OCB.

While OCB has been shown to be related to the variables of job satisfaction,

personality, affective disposition, and interpersonal skills, an important question that

needs to be considered is how are all of these variables related? Campbell’s (1990) job

performance model and the work of Campbell et al. (1993) provide some structure to

this issue. Campbell’s framework proposes the existence of multiple determinants of job

performance that can be applied to all types of performance, including OCB. To build a

prediction model, it was necessary to consider the aspects of Campbell’s model that

were most relevant to OCB. This prompted the requirement to define OCB and

understand its structure. Through the understanding of the OCB construct, the possible

role of antecedents could be considered.


The variables of motivation and ability form a common framework in

performance prediction models (e.g. Campbell, 1990). As previously discussed,

Campbell’s model describes three determinants of job performance, namely: declarative

knowledge, procedural knowledge and skills, and motivation. Although some individual

differences (e.g. personality) are associated with all three determinants (Campbell,

1990), declarative knowledge and procedural knowledge and skills are specifically

related to ability. Alternatively, attitudinal variables have been observed to influence the

motivation of the individual. This structure is consistent with the hierarchical influence

of personality, as described by Eysenck and Eysenck (1985). The current study

hypothesises that OCB would be predicted by ability and motivation and the structure of

the relationships are shown at Figure 3.9; as suggested by a modified model version by

Organ and Ryan (1995), taking into account the model as presented by Motowidlo et al.

(1997).

JOB ATTITUDES

DISPOSITION/PERSONALITY

FFM

Job Satisfaction

Positive Affect

KNOWLEDGE

SKILLS

ABILITY

x

DISPOSITIONAL

INCENTIVES

FFM



ORGANISATIONAL

CITIZENSHIP

BEHAVIOR

MEDIATOR

Motivation (Job Involvement)

Figure 3.9

Organisational citizenship behaviour prediction framework, as applied to the work of Organ

and Ryan (1995)


Using the framework presented in Figure 3.9, a hypothesised model was

developed using OCB antecedent variables that are related to either ability, motivation,

or both. Campbell’s (1990) performance model suggests the ability section of the model

includes variables that consist of declarative knowledge and procedural knowledge to

predict OCB. While traditionally ability has been related to task performance and more

specifically general mental ability, it has also been shown to be related to the procedural

skills/knowledge of self-management and interpersonal skills (Campbell et al., 1993).

The main OCB ability-related variable is interpersonal skills, as suggested by Ferris et

al. (2001), Campbell et al. (1993), and Motowidlo et al. (1997). As general mental

ability has been shown to be a cognitive predictor of interpersonal skills (see Ferris et

al., 2001), the model includes a pathway from general mental ability to interpersonal

skills. Additionally, the personality factors of conscientiousness, extraversion and

agreeableness are hypothesised to be predictors of interpersonal skills, based on the

results of Halfhill et al. (2008), Mount et al. (1998a), Neuman et al. (1999), Neuman

and Wright (1999), Tett et al. (1991), and Witt et al. (2002). A summary of these

relationships was used to develop a complex model to simultaneously assess the

prediction of OCB; shown at Figure 3.10.


To develop the section of the section of the model that was based on

motivational aspects, the current study uses the findings of previous correlational and

causal pathway research predicting OCB through job involvement, job satisfaction, and

related dispositional variables that were summarised in Figure 3.6 and reproduced at

Figure 3.11. Specifically, the previously established relationships, as discussed earlier in

this chapter are: extraversion to positive affect; positive affect to job involvement;

positive affect to job satisfaction; conscientiousness to positive affect; and

conscientiousness to job involvement.



Agreeableness

Conscientiousness

Extraversion

Figure 3.10

Initial organisational citizenship behaviour prediction framework for ability, as measured

through interpersonal skills.


Job

Satisfaction

Extraversion

Conscientiousness

Positive

Affect

Job Involvement

Combining the ability and motivation models at Figures 3.10 and 3.11 produces

the hypothesised OCB prediction model, which is shown in Figure 3.12. This model

aims to provide greater understanding of the complexity of OCB through the

simultaneous measurement of these important interacting relationships. Notably, the

personality traits of conscientiousness and extraversion are included in the prediction

pathways for both ability and motivation.

An additional benefit of the study design is that OCB and interpersonal skills

will be measured using a 360 degree feedback format. Specifically, these two variables

will use a self-report in addition to feedback reports from a supervisor, peer and

subordinate for each Petty Officer. This approach is discussed in the following chapter

and was taken to reduce the effects of self-report bias and common method variance.

The Chapter 5 will discuss the methodology of this study, while Chapter 6 will conduct

the analysis of the hypothesised model. Finally, Chapter 7 will discuss the results of the

analysis.

Figure 3.11

Initial organisational citizenship behaviour prediction framework for job involvement, as

measured personality, affective disposition and job satisfaction.


Job

Satisfaction

Extraversion

Conscientiousness

Positive

Affect



OCB

Job Involvement

Agreeableness

Figure 3.12

Hypothesised Model: Predicting organisational citizenship behaviour through

interpersonal skills and job involvement pathways. Interpersonal skills pathway involved:

general mental ability, extraversion, conscientiousness, and agreeableness. Job

involvement pathway involved: extraversion, conscientiousness, positive affect, and job

satisfaction.


CHAPTER FOUR – THREE HUNDRED AND SIXTY DEGREE FEEDBACK

Three hundred and sixty degree feedback has become an established method of

measuring behaviour within the domain of Industrial-Organisational Psychology. The

strategy extended the use of traditional forms of behavioural assessment (supervisor

reports and self-review) to having multiple observers from different organisational

levels rate the performance of a single individual. This form of performance feedback is

based on the construct of a circle, as data are collected from all source levels

surrounding the targeted individual. Bracken, Timmreck, and Church (2001) note that

the targeted employee receives behavioural ratings from supervisors, peers,

subordinates, and occasionally from clients and customers. A diagrammatic view of 360

degree feedback is demonstrated at Figure 4.1.

Reviewed

Individual

Self-Feedback

Feedback from

Supervisor

Top-down Feedback

Feedback

from Peers

Same Levels

Feedback

Feedback

from

Customers

and Clients

External

Feedback

Feedback from

Subordinates

Bottom-up

Feedback

Figure 4.1

360 degree feedback: The circle of feedback


Interestingly, 360 degree feedback was developed in a practical work

environment as part of performance evaluations and personnel development. Payne

(1998) stated the aim of this technique was twofold: to increase the objectivity of the

performance measurement and also to increase the employee’s self-awareness of how

their behaviour was interpreted by others. Bracken et al. (2001) note that this form of

performance assessment is also known by many alternative names, including: multi-

rater assessment, multi-source assessment, multi-source feedback, upward feedback,

and full circle feedback. For the purpose of the current study, all of these terms will be

referred to as 360 degree feedback.

This chapter briefly discusses the historical development of the 360 degree

feedback strategy. There is also consideration of the methodological and analysis issues

when 360 degree feedback is used as a performance development tool. Finally, these

issues will be discussed within the context of behavioural research and more

specifically the application of a 360 degree methodology in OCB research.

The Development of 360 Degree Feedback in the Working Environment

While the popularity of 360 degree feedback has grown dramatically over the

past three decades, Hedge, Borman, and Birkeland (2001) note that its origin was closer

to the beginning of last century. They argue that early in the twentieth century, the

relatively new profession of Industrial-Organisational Psychology was called upon to

create effective and efficient methods of recruiting and developing the increasing

number of workers that had moved to the cities following the industrial revolution. The

broad construct of performance appraisal was one of the foundations that these early

practitioners utilised to achieve this task.


Supervisor’s ratings were generally considered to be the accepted standard of

performance appraisal in the early twentieth century (Hedge et al., 2001). However,

even during those early days of performance appraisal, there was also an

acknowledgement of the usefulness of feedback from multiple sources to address

specific workplace issues. Shelton (1920) is recognised for using this strategy when he

conducted an investigation on mutual rating. He worked for an organisation that used

supervisor ratings within the performance appraisal system. However, it was identified

that inaccuracy was occurring during performance appraisals. To overcome the

inaccuracy issue, Shelton trialled a mutual rating strategy within one of the

organisation’s workgroups. Every employee in the targeted workgroup received one

supervisor rating in addition to ratings from peers and subordinates. As such, Shelton

employed a 360 degree feedback strategy, albeit not labelled so at the time. The ratings

occurred in a secret ballot and according to Shelton, provided an organisational

perspective of the individual. He considered that this information would be useful for

making promotion decisions so long as the process was fair, open and above board,

which he noted was not always easily achieved.

Further work that utilised an alternative source of feedback from the supervisor

report was conducted by Cook and Manson (1926). Their work aimed to identify the

specific traits and abilities that would be required by a successful retail sales employee.

The identified traits and abilities would be utilised in strategies to select and develop

sales staff. The authors considered that the customers were the best source of feedback

regarding the selling experience and as such, were selected to provide feedback on

specific retail traits and abilities. Notably the amount of sales was considered only to be

an indirect measure of customer satisfaction. Results supported the use of the customer

feedback as a valid method of performance appraisal of the identified traits.


Peer raters have also been observed to provide useful behavioural feedback in

performance assessments. An example of this approach is provided by Williams and

Leavitt (1947) who utilised peer reviews to evaluate the officers in the United States

Marine Corps. Based on their results, Williams and Leavitt concluded that peer reviews

were more valid at predicting the success of officers under training over several

objective tests; while also being a more valid predictor of future performance, as

compared to supervisor ratings. They suggested that peers were able to provide more

accurate observations as they spent more time with the targeted individual and co-

existed in realistic social situations. Peers directly interacted with the assessed

individual and were able to evaluate the social-dominance behaviour of that employee.

Importantly, the early work of Shelton (1920), Cook and Manson (1926), and

Williams and Leavitt (1947) supported the use of different rating sources as a data

collection strategy for performance appraisal. Interest in the technique continued to

grow and spread throughout private industry during the 1950 and 1960s (Hedge et al.,

2001). However, the process was initially very costly, as it was extremely time-

consuming and required a large personnel effort. In the 1980s and 1990s, 360 degree

feedback became widespread when efficiencies were made to the strategy’s

implementation in addition to employees and Human Resource practitioners becoming

more accepting of the process (Luthans & Farner, 2002). Notably, Luthans and Farner

argued that there was organisational change during this period in which individuals

could no longer expect to work for the same organisation throughout their career, as was

previously the case. As employees increasingly changed organisations, they became

more responsible for their own development and they actively sought increased

involvement in the performance feedback and development process. Also during this

period, job complexity increased and the traditional supervisor report had become


insufficient to provide adequate feedback to improve performance. Luthans and Farner

(2002) suggested that many companies collected the additional performance feedback

through the process of 360 degree feedback.

The Technique of 360 Degree Feedback to Develop Personnel

The methodology of 360 degree feedback has a long history based on the

broader construct of job performance appraisal. It is a core human development strategy

within the workplace (Craig & Hannum, 2006) and most commonly utilised for the

development of managers, in addition to counseling, training, and the validation of

personnel selection processes (Arnold, Cooper, & Robertson, 1998). More

controversially, 360 degree feedback has been used for administrative purposes,

including promotion and salary. However, this usage is debated due to a decreased

developmental benefit to the employee when the strategy is used simultaneously for

both developmental and administrative purposes; and also the difficulty of interpreting

feedback from different raters (Craig & Hannum, 2006).

Traditionally, organisations have relied upon the performance appraisal by the

supervisor to produce administrative decisions, including: development plans,

promotion, and future work objectives (Morgeson, Mumford, & Campion, 2005). Prior

to 360 degree feedback, employees could use the performance feedback to assess their

strengths and opportunities for improvement. In addition, the employees could use self-

reflection to improve their performance. Self-assessment provides a single point of

feedback and the accuracy is based on the individual’s self-awareness (Fletcher &

Baldry, 2000).


In contrast to a performance appraisal that is conducted for administrative

purposes, the main aim of 360 degree feedback is to develop the individual (Borman,

1997). Multiple observers provide additional information to the individual on how

his/her behaviour is viewed by other personnel across different levels of the

organisation. The characteristic of receiving performance feedback from multiple

sources to develop performance is considered to be an advantage. As such, Borman

argued that employees became more accepting of this strategy as the consequences were

considered to directly benefit the targeted employee. In time, improved performance

could enhance the opportunity for positive administrative consequences, including

promotion.

Rating Source Similarity

While the 360 degree feedback strategy has been widely accepted by industry,

there has been continued research interest surrounding the similarity of ratings between

different observers. The majority of this research considered initially if self-report

scores were similar to data collected from other sources. Harris and Schaubroeck (1988)

are recognised for conducting a pivotal meta-analysis to investigate the similarity of the

rating scores between self-reports and supervisors, self-reports and peers, and also

supervisors and peers. Using an extensive literature review, the authors summarised

three categories to explain why differences would be expected between different raters.

The categories were egocentric bias, differences in organisational levels, and

observational opportunities.

Harris and Schaubroeck (1988) argued that three forms of egocentric bias could

affect the size of the correlation between the raters. The first assumed that the self-

reporter would use a defensive strategy of biasing the score to enhance the performance


evaluation. The biased evaluation would create a restricted range, thereby reducing the

correlation between self-rating and other raters. Conversely, the authors did not expect

the existence of a range restriction for other raters. The second form of egocentric bias

involved the existence of a moderator variable. They provided the example of self-

esteem: individuals with high self-esteem may be more likely to inflate the score, while

individuals with low self-esteem would not. As such, correcting for range restriction

may not assist this form of bias unless the moderating variable was considered. The

third form of egocentric bias proposed by Harris and Schaubroeck was based on

attribution theory. The authors suggested that the difference between self and other

raters existed due to self-raters attributing good performance to their own effort and bad

performance to situational factors, while the reverse was true for other raters. As such,

the ratings would vary between self and other raters while being similar between other

raters. Based on the three egocentric bias effects, Harris and Schaubroeck suggested that

there would be a mean difference between self and other raters, with higher scores

provided by self-raters.

The second cause of difference between rater scores, as proposed by Harris and

Schaubroeck (1988), was due to the difference of the raters’ organisational level. The

authors suggested that two versions of this bias occurred. The first proposed that the

different levels of raters placed dissimilar weighting on each of the performance

dimensions. Secondly, performance was defined and measured differently between the

levels. Both of these versions hypothesised that a difference would exist between the

rater levels for the overall performance scores, while there would be no difference in

scores within the same level of rater.

Harris and Schaubroeck (1988) suggested the third explanation for the difference

between the raters was produced through different observational opportunities that were


afforded to each rater level. Specifically, peers would be more likely to view the rated

employee at more revealing times than the supervisor. As such, the strength of the

correlation between peers and supervisors would be reduced, as the supervisor did not

observe the same set of behaviours that were observed by the peer. The authors noted

that the difference in observational opportunity did not suggest a difference in

correlation size between self and other raters, similar to the second bias of

organisational differences. Notably, their assumption was in contrast to the findings of a

later study conducted by Fletcher and Baldry (2000), which suggested a difference of

rating scores between self and others due to self-ratings being based on self-awareness,

while other raters score the individual on external observations.

Harris and Schaubroeck (1988) addressed three questions in their study of the

similarity of rater scores. Firstly, what were the average correlations and difference

between each pair of raters and could any observed difference be explained by

egocentric bias, organisational level differences, and/or observational opportunities?

Secondly, could the differences between the pairs of raters be caused by statistical

issues, for example sampling error or restriction of range? Finally, once the statistical

issues were controlled for, were moderating variables able to explain the variance?

To assess their hypotheses, Harris and Schaubroeck (1988) analysed 36 self-

supervisor correlations, 23 peer-supervisor correlations, and 11 self-peer correlations

from studies that were published between 1956 and 1986. The authors reported a

correlation of 0.62 for peer-supervisor, which was higher than the correlations that were

observed between self-supervisor (ρ=0.35) and self-peer (ρ=0.36). As the correlations

between self and supervisors and self and peers were similar, no support was observed

using congruence-r analysis for the existence of differences across organisational levels.

Additional analysis was conducted to control for sampling error, measurement error,


and range restriction to assess if statistical issues caused a difference between the rater

scores. However, no reduction was observed between the correlations when these

statistical aspects were controlled for. Harris and Schaubroeck suggested this result did

not support the defensive bias strategy, however it was consistent with the egocentric

bias explanation of attribution theory – other rater agreement was higher than self-other

raters correlations. Notably, the higher correlation between ratings from supervisors and

peers could also be explained by Fletcher and Bawdy (2000) discussion of self-

awareness. Specifically, both supervisors and peers rated the individual using external

observations, creating similarity in response, while the self-report used internal

awareness, creating a difference between self-report and the ratings of others.

Harris and Schaubroeck (1988) noted that there was more than 25 percent of

unexplained variance in all three rater relationships. As such, they considered the

moderating variables of rating format (dimensional versus global), rating scale (trait

versus behavioural), and job type (managerial/professional versus blue-collar/service).

Results suggested that job type was the main moderator of the relationships.

Specifically, correlations for self-raters and both supervisors and peers were higher for

blue-collar/service jobs than managerial/professional positions. Notably, the

correlations were similar for peer-supervisor for both job types. The authors suggested

this result further supported that other raters had more agreement between their ratings

than with the self-rating, even in different job types.

The final analysis conducted by Harris and Schaubroeck (1988) was to compare

mean difference scores between the raters. Their analysis included 18 self-supervisor,

four self-peer, and seven peer-supervisor performance comparisons. While the authors

suggested the size and magnitude of the difference scores between self and supervisors

supported the egocentric-bias explanation, it is noted that the differences in the means


were non-significant. As such, no support was observed to suggest any significant

difference between raters’ means using this analysis.

The work by Harris and Schaubroeck (1988) is recognised for identifying

similarity and differences in rating patterns for raters during the process of 360 degree

feedback. Moreover, they considered some of the explanations as to why the similarities

and differences existed between the levels of raters. Notably, their work did not include

ratings from subordinates. As such, Conway and Huffcutt (1997) built on the work of

Harris and Schaubroeck to investigate the inter-rater reliabilities and correlations of

subordinates as a rating source, in addition to self, supervisor, and peer ratings. Their

meta-analysis consisted of 159 studies and compared the rating source to both job level

and job dimension. Results for all jobs indicated similar correlations between

subordinates and supervisors (r=0.22), subordinates and peers (r=0.22), and a lower

correlation with subordinates and self-reports (r=0.14). The correlation between

supervisor and self-rating was also 0.22, while the peer-self correlation was slightly

lower (r=0.19). Consistent with the research findings of Harris and Schaubroeck, the

strongest correlation was observed between supervisor and peer (r=0.34), while the

correlations between supervisors and peers were higher for non-managerial and lower

complexity jobs. Across the raters, higher between-rater correlations were observed for

the interpersonal dimension. The authors suggested that this result may indicate that

interpersonal behaviours were observed by all sources. In comparison, task behaviours

were more observable by supervisors, as compared to peers and subordinates.

The study findings of Harris and Schaubroeck (1988) and Conway and Huffcutt

(1997) suggest that the correlations between self and other rating sources are small to

moderate. Stronger correlations existed between other ratings, with the highest

correlation occurring between supervisors and peers. These studies contributed to the


identification of correlational patterns between the different levels of sources, while also

providing some evidence as to why differences occurred between raters. However, these

previous studies had not addressed if the difference in rating scores was due to the raters

working at specific organisational levels or the raters just being different people and

providing individual feedback. Mount, Judge, Scullen, Sytsma, and Hezlett (1998b)

argued that it was important to consider this aspect and they conducted a large-scale

study using a 360 degree approach that collected data from a sample of 2350 managers.

The notable methodological difference in their study was the collection of two feedback

scores from each rating level. This contrasts the standard collection strategy of one

report for each rating level. Specifically, each manager provided a self-appraisal, in

addition to receiving feedback from two supervisors, two peers, and two subordinates.

The design of the study enabled the researchers to investigate if the method effects in

the 360 degree feedback strategy were associated with the rater level, or each rater, or

both.

In their study, Mount et al. (1998b) tested five possible models that literature

had suggested could account for the difference in rated performance during the process

of 360 degree feedback. The five hypothesised models consisted of various

combinations of rating methods (raters) and traits (managerial skills). The managerial

skills of managers were assessed by the raters on the skill areas of administration,

human relations, and technical skills, and motivation. The first model considered only

traits and was based on a three-factor structure of the managerial skills. Models Two

and Three were seven-factor models. The seven factors of Model Two did not contain

any traits but rather included seven rating methods factors – two for bosses, two for

peers, two for subordinates, and one for self. In contrast, Model Three consisted of four

rating method factors (one for self, one for supervisors, one for peers, and one for


subordinates) and the three trait factors. Model Four contained a nine-factor structure –

six rating methods (one for self, one for supervisors, two for peers, and two for

subordinates) and the three traits factors. Finally, Model Five hypothesised a 10-factor

structure – seven rating sources (one for self, two for supervisors, two for peers, and

two for subordinates) and the three traits factors.

Following confirmatory factor analysis of each of the five models, Mount et al.

(1998b) suggested the results provided some support for both the nine-factor model

(χ2=1113.09 with 150 df; RMR=0.13; GFI=0.96; NFI=0.97; NNFI=0.96; CFI=0.97;

RFI=0.96) and the 10-factor model (χ2=333.95 with 144 df; RMR=0.02; GFI=0.99;

NFI=0.99; NNFI=0.99; CFI=0.99; RFI=0.99). Notably, the score for the Root Mean-

Square Residual (RMR) for the nine-factor model was in excess of the recommended

maximum level of 0.05 to indicate a good fit (Blunch, 2008). This result was not

discussed by the authors, however they did conduct additional testing and concluded

that the 10-factor model provided the best fit to the data. As discussed, this model

included the seven rating methods (one self, two supervisors, two peers, two

subordinates) and three trait factors (human relations, administration, technical) on the

manager being rated.

The support of the 10-factor model suggested that the method effects that

resulted from individual raters could be identified and contributed to the variance in

performance ratings more than the organisational level of the rater. Specifically, the

authors argued that the difference between raters’ scores actually occurred due to the

raters being different rather than working in different levels of the organisation.

Specifically, each rater possessed different traits, skills, and experience that produced a

different rating from other raters. Consequently, Mount et al. (1998b) considered that

360 degree feedback provided a more detailed appraisal of performance, as each


reporter provided a unique perspective, regardless of the organisational level that they

worked in. Moreover, they suggested that when the 360 degree feedback process is

utilised to assist in the development of the manager, the unique rater perspective should

be provided to the manager as separate feedback, rather than aggregated, as the

aggregation may hide some of the unique feedback. As such, if a score aggregation

occurred, the construct validity of the feedback would be reduced for the purpose of

individual feedback. Interestingly, while Mount et al. argued against the aggregation of

feedback due to a reduction of construct validity, they noted that rating reliability

improved when the scores were aggregated.

Researchers have continued to be interested in the degree of score similarity of

360 degree feedback raters. However, the majority of rater similarity studies have

focused on correlational analysis, which is termed congruence-r (Warr & Bourne,

2000). In addition to correlational analysis, Warr and Bourne (2000) suggest that

researchers should assess the mean difference between raters, which they termed

congruence-d. Congruence-d is defined as the closeness or difference of the judgments

that are assessed on either single or multiple performance dimensions. Warr and Bourne

(1999) suggest that congruence-d equates to effect size and it is obtained by subtracting

the average score by other raters for each rated item from the self-rated score, divided

by the pooled standard deviation of those scores. They stated that effect size is useful as

a standardised measure that can be compared across different studies. Additionally,

effect size captures the magnitude and direction of the difference in the judgments. A

positive congruence-d indicates over-rating by the self-rater or under-rating by other

raters, or a combination of both. Conversely, if congruence-d is negative, the ratings by

others are inflated, or the self-ratings are under-rated, or a combination of both. Warr

and Bourne (2000) suggested that different rating patterns for congruence-r and


congruence-d would be expected. Notably, the study conducted by Harris and

Schaubroeck (1988) assessed both forms of between rater agreement.

Warr and Bourne (2000) argued that as the majority of 360 degree feedback

studies had investigated the congruence-r of rating similarity rather than congruence-d,

further research was required. Specifically, they investigated if congruence-r and

congruence-d would be statistically independent with aspects of content rating. While

Harris and Schaubroeck (1988) had investigated aspects concerning the rater, Warr and

Bourne focused on behavioural aspects. The authors conducted a study to investigate

the association between the two forms of congruence with three forms of content rating

– the observability of the behaviour, the desirability of the behaviour, and the technical

versus interpersonal component of the behaviour. To investigate the relationships, Warr

and Bourne (2000) analysed 360 degree feedback data collected on 247 managers from

the United Kingdom. Ratings were collected for each manager from one self-report, one

supervisor report, two peer reports, and two subordinate reports. Notably the reliability

of the peer and subordinate were improved due to averaging the ratings across two

scores, while the self-report score and supervisor score were single scores for each

manager being assessed.

The initial analysis conducted by Warr and Bourne (2000) investigated the

relationship between congruence-r and congruence-d, which was followed by an

analysis of correlations between the pairs of raters for the two forms of congruence and

the 36 behaviours on the performance scale. The authors reported a mean

intercorrelation of -0.10 between congruence-r and congruence-d, which they argued

indicated that the two constructs were statistically independent. Average correlations

(congruence-r) between the raters and the assessed behaviours for the rater pairings of

self-subordinate (0.31) and self-peer (0.29) were higher than the observed relationship


of self-supervisor (0.21). In contrast, the average correlations between the effect sizes

(congruence-d) and the behaviours were not significant for the self to other raters’

comparisons, which was consistent with the earlier results of Harris and Schaubroeck

(1988).

The second phase of the study by Warr and Bourne (2000) investigated the

associations between the two forms of congruence and the three forms of content rating.

Within the congruence-r framework, Warr and Bourne noted the correlations between

self-supervisor (partial r=0.31) and self-peer (partial r=0.32) indicated that self-ratings

were higher than both supervisor and peer ratings when the behaviour was more easily

observed, once the desirability was controlled for. In contrast, when the behaviour was

more desirable, the negative correlations for self-supervisor (r=-0.51) and self-peer (r=-

0.34) indicated that both levels of rater scored the individual lower than the self-rating.

Notably, the self-subordinate correlation was not significant for both observability and

desirability of behaviour. When the observability of behaviour was analysed within the

congruence-d (effect sizes) framework, the observability did not produce a difference

between the raters, consistent to results of Warr and Bourne (1999). Similarly, the

desirability of behaviour failed to generate a difference for self-supervisors and self-

peers correlations. In contrast, desirable behaviour was observed to produce a

significant difference for the self-subordinate pairing. The difference score suggested

that the targeted manager over-inflated the score when the behaviour was desirable,

when compared to the rating by the subordinate.

Warr and Bourne (2000) also assessed congruence-r and congruence-d for the

third form of content rating – type of behaviour; by investigating three forms of

behaviour: interpersonal skills, cognitive and proactive behaviours, and technical

behaviours. Analysis revealed similar correlations between self and other rater scores


across the three forms of behaviours. In contrast, analysis of variance analysis

(ANOVA) and separate t-tests, demonstrated that the type of behaviour influenced the

pattern of difference for each rating pair. Specifically, Warr and Bourne suggested that

over-rating occurred for self-reports on interpersonal skills and others raters for

technical skills; while underrating occurred for self-reports on technical skills and other

raters for interpersonal skills. Furthermore, in a review of the mean ratings from all

raters, Warr and Bourne indicated that both the underrating and overrating of scores

across the different behaviours contributed to the differential over-rating.

The results from Warr and Bourne (2000) in addition to those from Harris and

Schaubroeck (1988), Conway and Huffcutt (1997), and Mount et al., (1998b) suggest

that patterns of similarities occurred between the raters of 360 degree feedback.

Specifically, these researchers observed small to moderate correlations between the

rating sources of 360 degree feedback. The strongest correlation was observed between

peers and supervisors. Moreover, Warr and Bourne (2000) observed opposing inflation

rating patterns from self-raters and others for interpersonal skills and technical skills.

Importantly, the observations from all raters were considered valuable to provide unique

feedback to develop the individual (Mount et al., 1998b). As such, the use of 360 degree

feedback has continued to be an important strategy to develop personnel in the

workplace (Kline & Sulsky, 2009).

OCB Research and 360 Degree Feedback

In addition to the use of 360 degree feedback as a personnel development tool,

researchers have begun to recognise the benefits of this strategy to measure behaviour

(Hedge et al., 2001). As such, this core human resource strategy has also been

increasingly used within behavioural research, particularly for skills that consist of


interpersonal aspects. However, there are issues that need to be considered when this

strategy is adopted for the study of many individuals, in comparison to providing

individual feedback to employees. Consequently, the understanding gained through

personnel development studies was reviewed to determine if it remained relevant for

behavioural research, including OCB research.

Notably, Organ et al. (2006) recommended that OCB research should

incorporate 360 degree feedback as part of the design methodology. The two main

reasons for this suggestion were that different raters may observe different forms of

OCB (Borman, 1997; Mount et al., 1998b); and secondly, utilising other raters in the

study design could address common method variance that occurs when using self-report

only, as discussed by Podsakoff et al. (2003) in Chapter 3. Organ et al. argue that each

rating source does not possess the same behavioural knowledge or motivation to report

on the individual. As both aspects are required to produce accurate ratings, it is assumed

that all rating sources for OCB contain some form of bias and subsequently, some levels

of inaccuracy.

Rating bias is one of the methodological issues that has possible implications for

behavioural research. Specifically, the culmination of the findings from Harris and

Schaubroeck (1988), Conway and Huffcutt (1997), Mount et al., (1998b), and Warr and

Bourne (2000) suggests that no rating source provides exactly the same score as another

rating source, either within or across rater levels. The difference of rater scores may

exist due to the characteristics of the raters, the behaviours being evaluated, the level

and type of position being evaluated, or why the evaluation was occurring. In a review

of the substantial theoretical and empirical literature, Fleenor, Smither, Atwater,

Braddy, and Sturm (2010) summarise the main aspects that can impact the accuracy of

source input (Table 4.1), thereby producing a difference of scores between rating


sources. As this list demonstrates, there are many bias categories that can affect rater

scores and consequently, increase or decrease the difference between the rating sources

of 360 degree feedback.

Table 4.1

Summary of Bias Categories for Rating Sources (Fleenor et al., 2010)

Self-Report Bias Categories Other Report Bias Categories

Biographical Characteristics

Gender, age, position, race, and education.

Rater Characteristics

Ability, job experience and performance, personality,

mood, beliefs about human nature, attitudes,

organisational commitment, discomfort with appraisal,

organisational level

Personality and Individual

Characteristics

Personality factors, dominance, empathy, self-

esteem, narcissism, intelligence, self-monitoring,

locus of control, and depression.

Rater Motivation

Goals, politics, accountability, and the incentive to

provide an accurate rating.

Context

Culture, controllability of behaviour, purpose of

report, and the similarity between self and other

raters.

Contextual Factors

Situational strength, culture, difficulty of the rating

task, observation context versus rating context,

negative versus positive performance incidents,

primacy versus recency effects, anchoring effects,

proportion of women in the workgroup, performance of

ratee’s peers, task interdependence, rater training,

rating purpose, trust in the appraisal process, and

understanding of norms.

Job Relevant Experiences

Experience.

Rater-Ratee Interactions and Expectations

Leader-member exchange, ratee’s past performance,

prior commitment to the ratee, rater expectation about

the ratee, familiarity with ratee, ratee impression

management tactics, ratee self-appraisals, similarity to

the ratee, and rater affect towards the ratee.

Given the complexity demonstrated in the bias summary at Table 4.1, I argue

that it would be practically impossible to prevent the influence of all of these biases in a

research setting. However, from the perspective of OCB research, is the bias between

and within rating levels relevant? In other words, does it matter if bias causes a


difference between rater scores for OCB? Importantly, Allen, Barnard, Rush, and

Russell (2000) conducted a study to investigate if supervisors and subordinates would

observe different types of OCB to self-reports, and also if bias could explain the

difference between rater scores. They hypothesised that in addition to self-bias inflating

the self-rating, the self-rater would score them self higher on OCBs than the supervisor,

as the supervisor would have a lack of awareness of OCBs occurring within the

workplace. This could occur if supervisors did not notice the occurrence of the

behaviour or by not having sufficient knowledge of what OCB is. Moreover, these

researchers suggested that as OCB was discretionary, any form of this behaviour may

not be publicly rewarded to the same extent, if at all, as compared to traditional forms of

job behaviour, such as task performance. Interestingly, Allen et al. recognised that some

subjective behaviour has been previously measured for the individual’s performance

appraisal. An example of this type of behaviour is oral communication, which is

included in the construct of interpersonal skills. As this type of behaviour has already

been reviewed by supervisors in regular performance appraisals, employees would have

an understanding of what behaviour was being measured. Moreover, supervisors would

be expected to measure a previously measured behaviour more accurately than a

behaviour that was not normally assessed.

To assess their hypotheses, Allen et al. (2000) sampled 372 participants from the

United States, including 41 rated individuals, 154 supervisors, and 177 subordinates,

however, their study did not measure peer ratings. Notably, while not mentioning

congruence-r and congruence-d specifically, the authors conducted analyses to assess

correlations between rating sources and the difference between their means. The initial

analysis reviewed the correlations between the different rating sources for both an

aggregate measure of OCB and also the five OCB factors, as described by Organ


(1988). Congruence-r results indicated that OCB ratings provided by supervisors and

subordinates were more strongly related (r=0.48, p<0.01) than either self-supervisor

ratings or self-subordinate ratings, which were both non-significant. Allen et al. also

analysed the mean difference between the ratings from each of the difference sources.

The ANOVA for the aggregated OCB score indicated that the means varied

significantly between the rating sources. Further analysis was conducted to identify how

the variation existed between the rating sources. The Tukey’s HSD test demonstrated

that self-bias inflation was observed between self and subordinate ratings. However, no

inflation was observed between self and supervisor ratings. This analysis suggested that

self and supervisors rated OCBs higher than the subordinate, which was contrary to the

hypothesised scoring pattern. To explain the result, the authors suggested that for their

sample, the self-rater was employed as a manager, were more likely to hold similar

OCB expectations and performance rating experience to the supervisor. In contrast, the

subordinate would have less knowledge of OCB as compared to the self-rater and was

also considered to be less trained to be able to provide upward feedback.

The results of the study by Allen et al. (2000) indicated that differences existed

between the rating sources and that some of the differences were based on bias,

including self-bias and the experience of the raters. Moreover, they argued that some

OCB dimensions were more observable than others. For example, attending meetings

versus helping people may have a different level of observability in the workplace by

different levels of observers. Due to the differences of OCB observation, Allen et al.

suggested that a single rater lowered the reliability of the score. To improve the

reliability, they argued that when measuring OCB in behavioural research, the

researcher should aggregate the scores of all of the observers to increase the overall


reliability of the OCB score. The aggregated score would provide a more

comprehensive score of OCB and include all of the unique perspectives of each rater.

The score aggregation approach recommended by Allen et al. (2000) was in

contrast to the unique feedback method that was suggested by Mount et al. (1998b). The

difference in their approaches has occurred as their studies were conducted within

different frameworks. The study by Mount et al. was aimed to provide unique individual

feedback to employees to enhance performance. In contrast, Allen et al. was

investigating patterns of ratings across large rating groups. As such, individual feedback

was not important in the second case. In a behavioural research context that did not

provide individual feedback, score aggregation could actually enhance the investigation

of behavioural trends, as the aggregation of rater scores would provide a more

comprehensive score for the behaviour being measured. Organ et al. (2006) also

recommended that 360 degree feedback could help reduce the bias across all the rating

inputs as an aggregated score could compensate for each other. Furthermore, Mount et

al. argued that when the feedback scores were aggregated, the use of multiple sources

could increase the reliability of the overall behaviour score.

While it is important to note that the measurement of OCB increased reliability

when aggregating scores from multiple observers, it is also critical to consider the

validity of the data collection process for the results to be useful (Fletcher, Baldry &

Cunningham-Snell, 1998). In other words, the data collection process must be both

valid and reliable to collect useful information to analyse. I propose that for OCB

research, the validity of the data analysis is improved when the OCB score is

aggregated. The reasoning for this assumption is that when the score is aggregated from

three observer levels ratings, the score then contains the different forms of OCB

behaviours that are observed by the different raters, while also combining the different


motivation and bias of the raters (Organ et al., 2006). The aggregated OCB score then

compensates for each rating source’s bias and knowledge, while also increasing the

reliability estimate, as discussed by Mount et al. (1998b). Notably, the same logic of

score aggregation can be used for measuring interpersonal skills.

360 Degree Feedback Rater Analysis for the Current Study

The strategy of 360 degree feedback offers an exciting new methodology for

OCB researchers. The current research will collect data for both OCB and interpersonal

skills utilising 360 degree feedback to obtain comprehensive scores for these two

variables, while also addressing the issue of common method variance. This research

thereby provides an opportunity to investigate some of the 360 degree feedback

methodological issues raised in the previously discussed personnel development studies

(e.g. Warr & Bourne, 2000) and determine if similar rater source patterns are

reproduced for OCB research. Specifically, the current study investigates the 360 degree

feedback rater similarity patterns for the constructs of OCB and interpersonal skills.

The first analysis will assess the similarity between rating sources using

congruence-r analysis. It has been observed that stronger correlations have occurred

between the other raters of performance than between self and other raters. The

strongest observed correlation has existed between supervisors and peers, as shown by

Harris and Schaubroeck (1988), Conway and Huffcutt (1997), Mount et al. (1998b), and

Warr and Bourne (2000). Notably, while Allen et al. (2000) observed that supervisors

had the strongest relationship with subordinates for OCB ratings, they did not measure

peer ratings. I would expect similar rating patterns that had been observed in personnel

development research (e.g. Harris & Schaubroeck, 1988; Conway & Huffcutt, 1997;

Mount et al., 1998b; Warr & Bourne, 2000) for both OCB and interpersonal skills. As


such, the current study proposes the following relationship patterns between raters of

OCB and interpersonal skills:

Hypothesis 1: The strongest correlation for OCB scores will occur between supervisor

and peer ratings.

Hypothesis 2: The strongest correlation for interpersonal skills scores will occur

between supervisor and peer ratings.

In addition to congruence-r analysis, the current study investigates the similarity

of rating sources through the concept of congruence-d. As the study aims to replicate

past patterns for observable and desirable behaviour, as discussed by Warr and Bourne

(2000), it is important to determine if these two constructs are considered observable

and desirable. Organ et al. (2006) argued that OCBs are desirable but not compulsory,

while different types of OCBs are observed by different levels of observers. Notably,

Williams and Anderson (1991) observed the OCBs that are directed at the organisation

are observed more by supervisors, while individually orientated OCBs are observed

more by peers and subordinates. As such, not every OCB would be expected to be

observed regularly by all raters, however, OCB is generally considered to be an

observable pattern of behaviour. When considering interpersonal skills within the

desirability-observability framework, they are considered both desirable (Ferris et al.,

2001) and observable by all sources (Conway & Huffcutt, 1997). Moreover, Williams

and Leavitt (1947) argued that interpersonal skills are specifically observed by peers

and subordinates. As OCB and interpersonal skills are considered desirable and

observable, there would be an expected difference between the rating sources.

Moreover, these differences could be based on an internal versus external frame of


observation, as suggested by Fletcher and Bawdy (2000). Specifically, the self-report

scores would be expected to be different to the scores of the other raters, as self-report

scores are based on internal self-reflection; while the other observers base their scores

on external observation. As such, the two hypotheses relating to the difference between

the means are:

Hypothesis 3: There will be a greater difference between the OCB scores of self and

others than between the other source raters.

Hypothesis 4: There will be a greater difference between the interpersonal skills scores

of self and others than between the other source raters.


CHAPTER FIVE – METHODOLOGY

The main aim of the current research was to develop and test a model to predict

OCB in the workplace. To achieve this goal, 360 degree feedback was incorporated

within the research design to address the issue of common method variance (see

Podsakoff et al., 2003), as discussed in Chapter 3. The multiple data collection and

analysis strategy of 360 degree feedback is relatively new to OCB research (Organ et

al., 2006) and as such, the current study provided an opportunity to investigate the

relationships between the feedback sources on the variables of interpersonal skills and

OCB, prior to model testing.

A cross-sectional design was used to measure the most relevant individual

differences and behavioural variables. The Royal Australian Navy was chosen as the

organisation to test the hypothesised model as the population would meet the required

selection criteria, while also providing benefit to the Royal Australian Navy through

additional personnel research. Military populations can provide reasonable sample sizes

for similar job positions both in terms of seniority and role. Notably, the Royal

Australian Navy and more broadly, the Australian Defence Force has experienced

increasing research on the relationship between personality and training performance

outcomes. However, there has been minimal consideration of individual traits and

subsequent job performance, specifically OCB. Consequently, the current study can

assist the development of knowledge in this domain. As the current study was

conducted using a population from the Royal Australian Navy, approval was sought and

gained through the Charles Sturt University Ethics Committee, the Australian Defence

Human Research Ethics Committee, and the Chief of Navy through the Director


General Naval Personnel and Training. Additionally, research advice was sought from

the Defence Force Psychology Organisation.

The rank of Petty Officer was selected to be the investigated position. A

summary of the Royal Australian Navy (2011) rank structure is provided at Appendix

D. This position is held by uniformed senior sailors with substantial corporate and

professional knowledge. Petty Officers typically have supervisory duties who also have

supervisors, peers, and subordinates who may observe their workplace behaviour.

Importantly due to the seniority of the position, a Petty Officer is more likely to have

the opportunity to undertake OCB related work behaviours, as compared to junior

members who generally have less discretionary behaviour opportunities in their work

role (Organ et al., 2006). It is also argued that OCB competence is beneficial for this

particular rank to be effective members in their chosen specialisation.

During the time of data collection I was a serving member of the Royal

Australian Navy. However, I was not in the direct chain-of-command of the

participants. Consequently, I could not place undue pressure on participants to partake

in the study. Utilising an informed consent process, all participants were assured that

their contribution was on a volunteer basis and non-participation would not affect their

Naval careers. They were able to withdraw prior to the completion of the study without

penalty and the research report would be written without specific reference to

individuals. An example of the research information sheet and the informed consent

form for the Petty Officer is contained at Appendix E. Petty Officers were requested to

ask a supervisor, peer, and subordinate to complete a behavioural questionnaire of their

interpersonal skills and OCB performance. An example of the research information

sheet and informed consent form provided to each supervisor, peer, and subordinate is

contained at Appendix F.


Participants

Petty Officers from 27 job group categories from the Royal Australian Navy

were selected as the target sample. The categories that were represented by the Petty

Officer volunteers were: Aviation, Aviation Technical, Bosuns Mate, Clearance Diver,

Combat Information System, Communications, Combat Systems, Cook, Electronic

Technical, Electronic Warfare Analysis, Marine Technical, Medical and Dental, Mine

Warfare, Naval Police Coxswain, Physical Trainer, Steward, Stores Naval, Submariner

Categories, and Writer.

Participants volunteered from Her Majesty’s Australian Ships (HMAS) in the

Sydney, New South Wales, and Rockingham, Western Australia areas, in addition to

Royal Australian Navy Establishments across Australia. The Establishments and related

lodger units were HMAS CERBERUS (Crib Point, Victoria), HMAS KUTTABUL

(Garden Island, Potts Point, New South Wales), HMAS PENGUIN (Mosman, New

South Wales), HMAS STIRLING (Garden Island, Rockingham, Western Australia),

HMAS WATSON (Watson’s Bay, New South Wales), HMAS CRESWELL (Jervis

Bay, Australian Capital Territory), HMAS ALBATROSS (Nowra, New South Wales),

HMAS HARMAN (Canberra, Australian Capital Territory), and Russell Offices

(Canberra, Australian Capital Territory). Given the level of operational activity and

personnel leave requirements, it was not possible to use random sampling techniques.

Instead, available candidates were requested to attend programmed information session

times if they were interested in volunteering for the study.

While this study originally aimed to assess 400 Petty Officers as a representative

sample of the wider Petty Officer community in the Royal Australian Navy, a high

operational commitment was present at the time of data collection. As such, the initial


target could not be achieved. A total of 202 Petty Officers volunteered to participate in

the study and complete self-report measures of: OCB, personality, general mental

ability, interpersonal skills, job involvement, job satisfaction, and affective disposition.

One Petty Officer requested to withdraw from the study at the conclusion of self-report

data collection and the collected information was subsequently removed. Another did

not complete all of the self-report scales and the related data were also removed,

resulting in a data sample of 200 Petty Officers. The initial sample consisted of 170

male and 30 female Petty Officer volunteers, ranging in age from 26 to 57 years

(M=35.76, SD=5.71), with a minimum education standard of year 10. However, for the

sample of 200 Petty Officers, reports from all three observer sources were obtained for

147 participants. This final sample of 147 cases consisted of 126 males and 21 females,

and ranged in age from 26 to 57 years (M=36.26, SD=5.90).

Materials

The current study required measurement for the variables of: OCB, general

mental ability, interpersonal skills, job satisfaction, job involvement, positive affect, and

the personality traits of extraversion, conscientiousness, and agreeableness.

Standardised testing occurred for general mental ability, personality, and affective

disposition. While the other variables were combined into two behavioural measures –

one for Petty Officers (enclosed at Appendix G – standard format; Appendix H – colour

coded to identify the item variable type) and one for the observer response group of

supervisors, peers, and subordinates (enclosed at Appendix I). Notably, all model

variables were measured using subjective scales (self-report or reports from other

observers). These scales were considered superior to objective measures to assess these

constructs (Muckler & Seven, 1992), as previously discussed in Chapter 3.


Standardised Materials

The measurement of general mental intelligence and personality was conducted

in accordance with their standardised instruction manuals. The affective disposition

scale was also applied in accordance with the standard guidelines; with a minor change

made to the response format. This change will be further discussed in the following

section. The affective disposition scale is enclosed at Appendix J.

General Mental Intelligence. Raven’s Standard Progressive Matrices (SPM)

(Raven, Raven, & Court, 1998) was used to measure the level of general mental

intelligence of the Petty Officers. This test was developed to assess an individual’s non-

verbal reasoning ability and was first published in 1938, then revised in 1958. The test

has been used widely for psychological assessment in the areas of clinical, education,

personnel selection, and research. Based on Spearman’s model of general intelligence,

the Standard Progressive Matrices predicts ability through the assessment of making

sense out of confusion, while minimising cultural implications (Grubb et al., 2004). It is

considered one of the purest single measures of ‘g’ and consists of 60 items arranged in

five sets of 12. Each item is in the form of a pattern which has a missing piece. The

individual is requested to choose from a set of six or eight pattern alternatives that

would complete the item pattern. As the testing was conducted in a group setting, the

format of an un-timed test was utilised to ensure effective measurement of perception

and clear thinking. As Australian norms were not available for the age group of the

Petty Officers, the scores were standardised using 1986 percentile norms for adults in

the United States, which was considered the closest population data group. Reported

reliability for this measure ranges from 0.83 to 0.93 across the age group of the Petty

Officers (Raven et al., 1998).


Personality. The NEO PI-R (Costa & McCrae, 1992a) measures five personality

dimensions to provide an overview of the individual. Additionally, the scale assesses six

facets in each dimension to provide a more thorough investigation of the individual’s

personality. Consequently, this scale provides a comprehensive overview of emotional,

interpersonal, experiential, attitudinal, and motivational styles. The NEO PI-R contains

240 items with three validity checks and requires a Year 6 reading level. The response

format is a five-item Likert scale ranging from: 1 – strongly disagree; 2 – disagree; 3 –

neutral; 4 – agree; 5 – strongly agree. The inventory yields a profile of T-scores using

Australian adult norming studies and results are separately normed for men and women.

Although the full scale was completed by participants, only the scores for the model

variables of agreeableness, conscientiousness, and extraversion were analysed.

Reliability estimates gained were adequate for these three personality dimensions

(agreeableness: r=0.72; conscientiousness: r=0.86; and extraversion: r=0.81). Costa and

McCrae report previous reliability scores for agreeableness, conscientiousness, and

extraversion factors to be approximately 0.86, 0.90, and 0.89, respectively.

Positive Affect. The PANAS (Watson et al., 1988) was used to assess affective

disposition. The measure involves rating 20-items, 10 per dimension of positive and

negative affect. A general time frame was used with the time instructions stating you

generally feel this way, that is, how you feel on average, as this measure is sensitive to

mood fluctuations when short time frames are utilised. Watson and et al. argue that

scores gathered using a longer or more general time frame exhibit trait-like stability.

While data were collected for both affective disposition dimensions, only the positive

affect data were analysed. The negative affect variable was not included in the OCB

prediction model and as such the related data were not analysed. The response format

used was a 6-point scale ranging from: 1 – not at all; 2 – very slightly; 3 – a little; 4 –


moderately; 6 – quite a bit; and 6 – extremely. The original response format contains

five-points with the first point combining “not at all” and “very slightly”. To distinguish

between these two responses, the present study separated these choices into two

individual points creating six points in total. Internal consistency with a general

timeframe gained in the present study was 0.88 for positive affect, which is consistent

with the findings from Watson and colleagues.

Behavioural Scale Items

Two forms of the behavioural scales were produced. One form of the

behavioural scale was used to assess Petty Officers and the other version provided

feedback on the performance of the Petty Officer by observers. A common Likert scale

response format was used for the two behavioural scales, the responses being: 1 –

strongly disagree; 2 – disagree; 3 – slightly disagree; 4 – slightly agree; 5 – agree; 6 –

strongly agree. It is important to note that there was no neutral or central rating item for

undecided/unknown or not applicable. The neutral position can be useful for

respondents who prefer not to answer the question, do not know or understand the

question, or have a neutral position to the question. As the measure was questioning

volunteer respondents about personal attitudes and/or observable behaviours that were

desired in the workplace, as judged by Royal Australian Navy subject matter experts,

the scale design removed the possibility of central scoring bias. Respondents were

forced to decide if they agreed or disagreed and to what extent this occurred.

The self-report behavioural measure for the Petty Officers contained 89 items to

measure: OCB, interpersonal skills, job satisfaction, weighted job satisfaction, and job

involvement in a jumbled order. The order of these items was decided by placing all of

the items into a hat and then drawing them one at a time. The self-report behavioural


questionnaire was trialled on five non-population Royal Australian Navy personnel. The

trial indicated that self-report assessment including the scales for general mental ability,

personality, positive affect, and the behavioural scale required up to two hours per data

collection session.

The observer respondents were asked to complete a questionnaire containing 60

items that measured OCB and interpersonal skills. The observer behavioural measure

was also trialled by five non-population Royal Australian Navy personnel and took

approximately 15 to 30 minutes to complete.

There were no established measures that were considered to have all the

necessary attributes to measure job involvement, interpersonal skills, job satisfaction,

and weighted job satisfaction in the current study. Consequently, previously established

scales were modified to measure these variables. Additionally, the traditional scale to

measure OCB that was developed by Podsakoff et al. (1990) was reviewed for currency.

The following discussion provides a more detailed review of the behavioural items to

measure the variables of job involvement, job satisfaction, OCB, and interpersonal

skills.

Job Involvement. The construct of motivation is very broad and varies across the

theoretical perspective, as previously mentioned in Chapter 3. From a workplace

perspective, researchers have tended to measure motivation indirectly through variables

that form part of the larger motivation construct, such as job involvement. Job

involvement is based on personal values and attitudes and is conceptually different from

job satisfaction. Cook, Hepworth, Wall, and Warr (1981) note that the Job Involvement

Scale (JIS) (Lodahl & Kejnar, 1965) is frequently used as a measure of job motivation

aspects. It is a short and easily completed measure that has face validity. The JIS is a 20


item scale, seven of which are negatively phased. An additional item was included in

the scale to assess if the Petty Officer thought his/her Naval career was important. The

final 21 items are listed in Table 5.1. Twelve of the JIS items remained unchanged from

the original 1965, while eight items were slightly modified for currency and relevance

in a Navy working environment. For example, some of the changes included the word

he being replaced with s/he and the word depressed was replaced with upset. If an item

was altered, the original version is included in italics. The study produced a reliability

estimate of 0.79 for the job involvement items. Past studies have indicated reasonable

reliability ranging between 0.62 and 0.93 (Cook et al., 1981).

Table 5.1

Job Involvement Items

Item

1 My Naval career is very important to me.

2 I usually show up for work a little early, to get things ready.

3 The major satisfaction in my life comes from my job.

4 I enjoy thinking ahead to the next day’s work.

(Original version: Sometimes I lie awake at night thinking ahead to the next day’s work)

5 I avoid taking on extra duties and responsibilities at work. (R)

6 Quite often I feel like staying home or in my rack, instead of coming to work. (R)

(Original version: Quite often I feel like staying home from work instead of coming in)

7 To me, my work is only a small part of who I am. (R)

8 Some of the most important things that happen to me involve my work.

(Original version: The most important things that happen to me involve my work)

9 I really like to get things right at work.

(Original version: I am really a perfectionist about my work)

10 I feel upset when I fail at something connected with my job.

(Original version: I feel depressed when I fail at something connected with my job)

11 I have other activities more important than my work. (R)

12 I used to be more ambitious about my work than I am now. (R)

13 I am very personally involved in my work.

14 I would probably keep working in the Navy even if I didn’t need the money.

(Original version: I would probably keep working even if I didn’t need the money)

(continued next page)


Table 5.1 (continued)

Job Involvement Items

Item

15 You can measure a person pretty well by how good a job s/he does.

(Original version: You can measure a person pretty well by how good a job he does)

16 For me, mornings at work really fly by.

17 I live, eat, and breathe my job.

18 Sometimes I’d like to kick myself for the mistakes I make in my work.

19 Most things in my life are more important than work. (R)

20 I used to care more about my work, but now other things are more important. (R)

21 I’ll stay back to finish a job, without being asked by my supervisor, even if I’m not given

time in lieu. (Original version: I’ll stay overtime to finish a job, even if I'm not paid for it)

Items based on JIS as developed by (Lodahl & Kejnar, 1965).

Item 1 was added to the original JIS items.

Amended JIS items detail the original version.

(R) Items reversed scored.

Job Satisfaction and Weighted Job Satisfaction. The Short-Form Minnesota

Satisfaction Questionnaire (MSQ) (Weiss, Dawis, England, & Lofquist, 1967) was used

as the measure of job satisfaction. The short form, as listed at Table 5.2, consists of 20

items and was chosen over the long form version due to time consideration. To improve

the currency of the scale, four additional items were included that are relevant to job

satisfaction in the workplace of Royal Australian Navy personnel. The four questions

were identified in attitudinal research that was conducted for the Royal Australian Navy

(Johnson & Jacka, 2005; McKinnon, Montalban, Barton, & Gloyn, 2005; Power &

Nottage, 2005). The additional items are listed in Table 5.2 and considered the job

characteristics of work-life balance, required travel, mentoring opportunity, and the

quantity of work.

The resulting 24 job satisfaction items formed part of the self-report behavioural

scale and used the common scoring 6-point scoring scale. Some items were slightly

modified for use in a Royal Australian Navy population. For example the term Navy in

place of organisation or company, and personnel was used in place of the term men,


while the word always was removed from the items as this is an extreme term. All items

were modified from their original formats to ensure that they were grammatically

correct in relation to the response options. Specifically the terms: I am satisfied… or I

am not satisfied… were included at the beginning of each job satisfaction item. Cook et

al. (1981) notes that past reliability coefficients for the MSQ in its standard form have

generally been greater than 0.80, which is consistent with the reliability of 0.79 that was

achieved in this study.

In addition to overall job satisfaction, I wanted to be confident that the job

satisfaction scale items were actually important to the Petty Officers. To achieve this

assessment, I developed a scale to assess the importance of each job satisfaction item.

Based on the MSQ (Weiss et al., 1967) and the four additional job satisfaction items, a

24 item scale was produced requesting Petty Officers to rate how important each job

satisfaction aspect was to them. A comparison between the job satisfaction items and

the corresponding job satisfaction importance items is also contained at Table 5.2.

Reliability estimate for the job satisfaction importance scale was 0.76.

Table 5.2

Job Satisfaction and Corresponding Job Satisfaction Importance Items

Job Satisfaction Item Job Satisfaction Importance Item

1 I am satisfied with the way my job provides

for steady employment.

(Original version: The way my job provides for

steady employment)

1 Steady employment is important to me.

2 I am satisfied with the level of activity in the

workplace.

(Original version: Being able to be kept busy all

the time)

2 Being able to keep busy all the time is

not important to me. (R)

3 I am not satisfied with the chance to work

alone. (R)

(Original version: The chance to work alone on

the job)

3 The chance to work alone is important

to me.

(Continued next page)





4 I am satisfied with the chance to do different

things from time to time.

(Original version: The chance to do different

things from time to time)

4 The chance to do different things from

time to time is important to me.

5 I am satisfied with the chance to be

“somebody” in the wider community.

(Original version: The chance to be “somebody”

in the community)

5 The chance to be “somebody” in the

wider community is not important to

me. (R)

6 I am not satisfied with the way my boss

relates to his/her subordinates. (R)

(Original version: The way my boss handles his

men)

6 The way my boss relates to his/her

subordinates is important to me.

7 I am satisfied with the competence of my

supervisor in making decisions.

(Original version: The competence of my

supervisor in making decisions)

7 The competence of my supervisor in

making decisions is important to me.

8 I am satisfied with the balance between work

requirements and family life.

8 The balance between work

requirements and family life is not

important to me is important to me. (R)

9 I am not satisfied with the chance to do

things for other people. (R)

(Original version: The chance to do things for

other people)

9 The chance to do things for other people

is important to me.

10 I am satisfied with the chance to tell people

what to do.

(Original version: The chance to tell people what

to do)

10 The chance to tell people what to do is

important to me.

11 I am satisfied with the chance to do

something that makes use of my abilities.

(Original version: The chance to do something

that makes use of my abilities)

11 The chance to do something that makes

use of my abilities is not important to

me. (R)

12 I am not satisfied with the way Navy’s

policies are put into practice. (R)

(Original version: The way the company policies

are put into practice)

12 The way Navy’s policies are put into

practice is important to me.

13 I am satisfied with my pay for the amount

and quality of work I undertake.

(Original version: The pay and the amount of

work that I do)

13 My pay for the amount and quality of

work I undertake is important to me.






14 I am satisfied with the chances of

promotion.

(Original version: The chance for advancement

on this job)

14 The chances of promotion are not

important to me. (R)

15 I am satisfied with the level of freedom to

use my own initiative.

(Original version: The freedom to use my own

judgment)

15 The freedom to use my own initiative is

important to me.

16 I am not satisfied with the chance to try my

own methods of doing the job. (R)

(Original version: The chance to try my own

methods of doing the job)

16 The chance to try my own methods of

doing the job is important to me.

17 I am satisfied with the working conditions.

(Original version: The working conditions)

17 The working conditions are not

important to me. (R)

18 I am satisfied with the amount of mentoring

I receive during my career.

18 The amount of mentoring I receive

during my career is important to me.

19 I am not satisfied with the way my peers get

along with each other. (R)

(Original version: The way my co-workers get

along with each other)

19 The way my peers get along with each

other is not important to me. (R)

20 I am satisfied with the recognition I get for

doing a good job.

(Original version: The praise I get for doing a

good job)

20 The recognition I get for doing a good

job is important to me.

21 I am satisfied with being able to do things

that don’t go against my conscience.

(Original version: Being able to do things that

don’t go against my conscience)

21 Being able to do things that don’t go

against my conscience is important to

me.

22 I am satisfied with the amount of travel

undertaken.

22 The amount of travel undertaken is

important to me.

23 I am not satisfied with the amount of work I

have to finish in normal working hours. (R)

23 The amount of work I have to finish in

normal working hours is not important

to me. (R)

24 I am satisfied with the feeling of

accomplishment I get from the job.

(Original version: The feeling of accomplishment

I get from the job)

24 The feeling of accomplishment I get

from the job is important to me.

Job satisfaction items adapted from the MSQ developed by Weiss et al. (1967).

Items 8, 18, 22, 23 are the additional items that were included in the job satisfaction scale based on the

work in Royal Australian Navy attitudinal surveys (Johnson, & Jacka, 2005; McKinnon, Montalban,

Barton, & Gloyn, 2005; Power & Nottage, 2005)

(R) Reverse Scored


Each job satisfaction item score was multiplied with the corresponding job

importance scale item score to provide a weighted job satisfaction score. The weighted

job satisfaction score involved the level of satisfaction the Petty Officer had with an

aspect of his/her employment and if that aspect was important to the Petty Officer. Both

variables are included in the analysis in Chapter 6.

Organisational Citizenship Behaviour Items

The present study required two forms of OCB measurement – one self-report

and the other for the observers. The most common OCB measure was developed by

Podsakoff et al. (1990). The measure consists of 24 items that assess the five

dimensions of OCB, as proposed by Organ (1988). The following discussion details the

development of the observer and self-report forms of the scales.

Organisational Citizenship Behaviour – Observer Items. OCB is a group of

generic behaviours that are present in all organisations. The scale developed by

Podsakoff et al. (1990) has been widely used in OCB research. However, it is important

to highlight that this scale was developed in 1990. To ensure that it remained current for

today’s working environment in the Royal Australian Navy, additional OCB items were

considered. To identify possible additional OCB items for comparison with the

Podsakoff et al. (1990) scale, four further scales that researchers have also used were

reviewed. These scales in their original formats were:

1. Smith et al. (1983) OCB measure;

2. Motowidlo and Van Scotter (1994) contextual performance measure;

3. Hui, Law, and Chen (1999) OCBI and OCBO scale; and

4. Hannam and Jimmieson (2002) OCB scale.


The Podsakoff et al. (1990) OCB scale was based initially on the work

conducted by Smith et al. (1983) producing a duplication of items between these two

scales. Additionally, many of the items from the other three scales were created using

variations of the Podsakoff et al. (1990) scale items. Duplicated items across the

additional scales were deleted. Some of the selected items were rephrased for use in a

military environment. An additional 32 items, including one that was developed

specifically for the current study (willingness to ask people for help), were added to the

initial 24 items from Podsakoff et al. (1990), creating a trial OCB scale of 56 items.

This scale is contained at Appendix K and contained both positive and negative phrased

behaviours.

To determine if the trial OCB items were relevant for the Royal Australian

Navy, they were sent to subject matter experts for assessment. The most appropriate

group in the Royal Australian Navy to review the proposed items were Category

Sponsors. Category Sponsors manage the employment of specific Royal Australian

Navy job groups known as categories. Category Sponsors manage their category

through the provision of remuneration, competency development, and workforce

planning and structure policy (Department of Defence, 2001).

A total of 82 Category Sponsors and their staff rated the importance of the 56

proposed OCB items. An example for the research information sheet and informed

consent form provided to the Category Sponsors is enclosed at Appendix L.

Respondents were requested to identify if the item was an important behaviour for the

role of a Petty Officer in the Royal Australian Navy. A four-item response scale was

used, ranging from 1 – not important, 2 – slightly important, 3 – important, and 4 – very

important. Negatively phrased items were reverse scored to gain a comparative score

with positively phrased items. Thirty eight completed questionnaires were returned.


OCB items were included in the final observer behavioural scale if they scored an

average of more than 3.0 across the returned questionnaires from the Category Sponsors

(i.e. the behaviour was rated as important). In addition to the Podsakoff et al. (1990)

scale, a further 21 items were subsequently included creating 45 OCB items. The 45

OCB items are listed in Table 5.3 and include their source.

Table 5.3

Final 45 Organisational Citizenship Behaviour Items for Observer Assessment

Item Item Source

1 Helps other personnel who have heavy work loads. Podsakoff et al. (1990)

2 Is a classic “squeaky wheel” that always needs greasing. (R) Podsakoff et al. (1990)

3 Believes in giving an honest day’s work for an honest day’s pay. Podsakoff et al. (1990)

4 Spends a lot of time complaining about trivial matters. (R)

Podsakoff et al. (1990)

5 Tries to avoid creating problems for co-workers.


6 Keeps up and adjusts with changes in the organisation.


7 Tends to make “mountains out of molehills”. (R)


8 Considers the impact of his/her actions on co-workers.


9 Attends meetings that are not mandatory, but considered

important.


10 Is ready to lend a helping hand to those around him/her.


11 Attends functions that are not required, but help the Navy’s image.


12 Reads and keeps up with Navy’s announcements, policy, etc.


13 Helps others who have been absent.


14 Does not abuse the rights of others.


15 Willingly helps others who have work related problems.


16 Focuses on what’s wrong rather than the positive side. (R)


17 Takes steps to try to prevent problems with other workers.


18 Attendance at work is above the norm.


19 Always finds fault with what the Navy is doing. (R)


20 Is mindful of how his/her behaviour affects other personnel’s jobs.


21 Does not take extra breaks.


22 Complies with Navy’s rules even when supervisors are not

present.





Final 45 Organisational Citizenship Behaviour Items for Observer Assessment

Item Item Source

23 Helps orient new personnel even though it is not required.


24 Is an extremely conscientious worker.


25 Displays proper military appearance and bearing. Motowidlo and Van Scotter (1994)

26 Follows proper procedures and avoids unauthorised

shortcuts.

Motowidlo and Van Scotter (1994)

27 Pays close attention to important details. Motowidlo and Van Scotter (1994)

28 Supports the supervisor’s decisions. Motowidlo and Van Scotter (1994)

29 Renders proper military courtesy. Motowidlo and Van Scotter (1994)

30 Takes the initiative to solve a work problem.


31 Exercises personal discipline and self-control.


32 Is punctual, arriving at work on time and after breaks.

Smith et al. (1983)

33 Coasts towards the end of the day. (R)

Smith et al. (1983)

34 Spends a great deal of time in personal telephone

conversations. (R)

Smith et al. (1983)

35 Makes innovative suggestions to improve the quality of

the department.

Smith et al. (1983)

36 Shows concern and courtesy toward co-workers, even

when under stress.

Smith et al. (1983)

37 Often speaks ill of the supervisor or other personnel

behind their backs. (R)

Hui et al. (1999)

38 Takes credit, avoids blame, and fights fiercely for

personal gain.

Hui et al. (1999)

39 Uses Navy resources to do personal business. (R)

Hui et al. (1999)

40 Checks on co-workers who have been upset or struggling

with a problem.

Hannam and Jimmieson (2002)

41 Consults with co-workers when planning tasks that may

affect them.


42 Acquires knowledge and skills in new areas that

contribute to the Navy.


43 Takes care to conserve resources and look after

equipment.


44 Willing to ask for help from other personnel when

required.

Teal – created for present study

45 Uses illicit tactics to seek personal influence regardless of

interpersonal consequences.(R)

Hui et al. (1999)

Items 1-24 inclusive are items adapted from the Podsakoff et al. (1990) OCB scale

Items 25-45 inclusive are the remaining items from the other four scales (Hannam &

Jimmieson, 2002; Hui et al., 1999; Motowidlo & Van Scotter, 1994; Smith et al., 1983)

(R) Reversed scoring


The observer group was asked to respond to the 45 OCB items that were

included in the observer behavioural measure. To determine if the Podsakoff et al.

(1990) scale remained current for use in today’s workforce, the scores from the

additional 21 items were correlated with the scores from the Podsakoff et al. (1990)

scale. Correlations between the Podsakoff et al. scale items and the new OCB items

were assessed for the responses by supervisors, peers, and subordinates. Strong

relationships were observed between the Podsakoff et al items and the other items for

all observer ratings (supervisors: r=0.90; peers: r=0.84; and subordinates: r=0.89). As

the correlations ranged from 0.84 to 0.90 across all three observer groups, I considered

that the Podsakoff et al. (1990) scale was measuring OCB similarly to the 21 items

included via the Category Sponsors review. Following the results of this evaluation, the

Podsakoff et al. (1990) scale was considered sufficiently current to measure OCB in the

current study.

The current study also provided an opportunity to consider the similarity of

OCB scores provided by the different rating sources. Mount et al. (1998b) and Allen et

al. (2000) argue that the reliability of a scale is increased when the scores from the

different raters are aggregated. As such, the aggregated OCB score from the three

observers was used in the model testing. The aggregated score was based on the

Podsakoff et al. (1990) scale items from all three observer groups – supervisors, peers,

and subordinates. Reliability for the 24 items of the Podsakoff et al. (1990) scale using

the aggregated observer score across the 24 items was 0.95.

Organisational Citizenship Behaviour – Self-Report Items. The OCB items from

the observer measure could not be utilised by the Petty Officers as many of the items

were overtly preferred or non-preferred behaviour. It was unlikely that the Petty Officer

would be able to respond objectively to these items. It was decided to restrict the self-


report scale to five OCB items due to the high likelihood of self-report bias.

Specifically, as discussed in Chapter 4, self-raters would be expected to inflate their

scores based on social desirability and ego (e.g. Harris & Schaubroeck, 1988; Warr &

Bourne, 2000). The following five items were related to Organ’s (1988) five factor

OCB structure of altruism, generalised compliance, courtesy, sportsmanship, and civic

virtue, respectively:

I go out of my way to help other personnel;

I maintain a high level of effort at work even when my supervisor is not present;

I am courteous with other personnel to maintain working relationships;

I openly support Navy’s policy and practice’s even though I may not personally

agree with them; and

I volunteer for secondary duties and positions on committees that are not

mandatory but are important.

While data were collected from Petty Officers for a self-report OCB score, it

was not used for model testing but rather only in the correlational analysis with other

measured variables. Reliability alpha for the self-report form of these items was 0.56.

Interpersonal Skills – Observer and Self-Report Items. At the time of data

collection, there was no previously established interpersonal skills scale that was

considered relevant for the current study. A generic scale was developed using the

seven-item social skill scale of Ferris et al. (2001) as the basis. Five items from this

scale were selected. Ten additional items were included based on the work of Argyle

(1994), Hargie, Saunders, and Dickson (1994), and Hayes (2002). Items were

considered for inclusion based on the following behaviours: verbal and non-verbal


communication, written communication, interviewing or questioning, asserting and

influencing, handling feelings and emotions, handling conflict, handling diversity,

adaptability in social environments, helping and facilitation, empathy, listening, and

negotiating. A total of 15 items were identified to assess interpersonal skills.

The 15 items were worded to produce two forms of the scale – one for self-

report and one for observers. To analyse the reliability of both forms of the

interpersonal skills scale and the 15 items, Cronbach’s Alpha was calculated based on

the ratings from each respondent group (Petty Officer – self-report: α=0.75; supervisor

– observer: α=0.87; peer – observer: α=0.82; subordinate – observer: α=0.88). Items that

were negatively worded in the scale were reversed scored for the purpose of analysis.

The Likert scale ranged between one and six. All items were expected to score above

3.5, which would indicate the item was measuring interpersonal skills. Fourteen items

across the supervisors, peers, and subordinates scores achieved mean scores greater than

4.02. However, one item ranged from 2.29 to 2.36 across the three observer scores. The

lower scoring item was aimed to assess adjustability across situations and was phrased

in reverse score form. The item was worded for observers as “Keeps his/her behaviour

constant independent of the situation” and “I keep my behaviour constant independent

of the situation” for the self-report form of the item. This result suggested that the

adjustability did not measure interpersonal skills as expected.

To determine why the adjustability item did not measure interpersonal skills as

expected, further discussions were held with 10 Petty Officer and 10 observers selected

from the population. Notably, they reported some confusion when responding to this

item. It was theorised that a person high in interpersonal skills should be adaptive to the

environment and audience. In contrast, some respondents stated that in a military

environment, an individual was rated higher if they were predictable and unchanging,


particularly in emergency situations. The ambiguity contained in this item produced a

non-consistent result for the purpose of measuring the construct of interpersonal skills.

The item was subsequently removed from further analysis for both forms of the

interpersonal skills scale. The reliability estimates were reassessed for the remaining 14

items, which produced increased Cronbach’s Alpha for Petty Officers (α=0.80),

supervisors (α=0.90), peers (α=0.86), and subordinates (α=0.92) responses. This action

enabled acceptable reliability for both forms of the interpersonal skills scale.

Similar to the OCB self-report, self-report data collected for interpersonal skills

was used in the similarity analysis for 360 degree feedback. The detail of the similarity

analysis conducted for interpersonal skills raters is contained in Chapter 6. The self-

report data were not used in the model testing. Alternatively, the model testing used an

aggregated interpersonal skills score from the ratings scores of supervisors, peers, and

subordinates. Reliability for the 14 items using the aggregated observer score was 0.92.

The final 14 items for the measurement of interpersonal skills for both self-report and

observers are contained in Table 5.4.

Table 5.4

Interpersonal Skills Items for Self-Report and Observer Measures

Self-Report Item Observer Item

1 I express myself clearly to other

personnel.

1 Expresses him/herself clearly to other

personnel.

2 I find it difficult to put myself in the

position of others. (R)

2 Finds it difficult to put him/herself in the

position of others. (R)

3 I am good at sensing the hidden agendas

of others.

3 Good at sensing the hidden agendas of

others.

4 I find it easy to negotiate my position

with other personnel.

4 Finds it easy to negotiate his/her position

with other personnel.

5 I am very good at reading other

personnel’s body language.

5 Very good at reading other personnel’s

body language.

6 I find it difficult to handle diversity

within the workplace. (R)

6 Finds it difficult to handle diversity within

the workplace. (R)

(continued next page)


Table 5.4 (Continued)

Interpersonal Skills Items for Self-Report and Observer Measures

Self-Report Item Observer Item

7 I am keenly aware of how I am perceived

by others.

7 Keenly aware of how he/she is perceived

by others.

8 In social situations, I always know what

to say.

8 In social situations, he/she always seems

to know what to do and say.

9 I have effective listening skills. 9 Has effective listening skills.

10 My writing style is clear. 10 Has a clear writing style.

11 I find it difficult to adjust my

communication style across different

audiences. (R)

11 Finds it difficult to adjust his/her style of

communication across different

audiences. (R)

12 I create conflict with other personnel. (R) 12 Creates conflict with other personnel. (R)

13 My assistance is appreciated when I help

other personnel.

13 The assistance from the PO is appreciated

when he/she helps other personnel.

14 I influence other personnel to get them to

complete tasks.

14 Influences other personnel to get them to

complete tasks.

(R) Reversed scoring

Summary of Measures

The current study developed an OCB prediction model consisting of the

variables: OCB, interpersonal skills, job involvement, general mental ability, job

satisfaction, positive affect, and the personality factors of extraversion,

conscientiousness, and agreeableness. To measure these variables, standardised scales

were used where possible. Where this was not achievable, modifications were made to

existing scales or developed new scales. Table 5.5 lists the studied variables and

summarises how they were measured. Additionally, Table 5.5 states who completed the

scale, which was either self-report from the Petty Officer or the observer group

incorporating the supervisors, peers and subordinates. The OCB prediction model was

assessed using structural equation modeling techniques and will be discussed in Chapter

6.


Table 5.5

Summary of Measures Utilised for Data Collection

Variable Final Measure Self-

Report

Observer

Report

Standard or Modified Scale

OCB Modified versions of

the Podsakoff et al.

(1990) scale

Yes –

five

items

Yes – 24

items

Initially reviewed and compiled from

a range of measures. Only observer

data used for model testing.

Personality Revised NEO

Personality

Inventory (NEO PI-

R) (Costa &

McCrae, 1992a)

Yes No Standard as per the testing guidelines.

General

Mental

Intelligence

Raven’s Standard

Progressive Matrices

(Raven et al., 1998)

Yes No Standard as per the testing guidelines.

Positive

Affect

PANAS (Watson et

al., 1988)

Yes No Standard items with modified

response format and combined with

other behavioural measures. Only

utilise the Positive Affect scores for

data analysis.

Job

Involvement

JIS (Lodahl &

Kejnar, 1965)

Yes No Standard items with modified

response format and combined with

other behavioural measures.

Job

Satisfaction

Short-Form MSQ

(Weiss et al., 1967)

Yes No Amended items containing the

emphasis of the original items with

modified response format and

combined with other behavioural

measures. Four additional items from

Royal Australian Navy attitudinal

survey work.

Weighted Job

Satisfaction

New measure in

conjunction with the

MSQ.

Yes No Developed items to correspond with

the Standard items from MSQ with

modified response format and

combined with other behavioural

measures.

Interpersonal

Skills

New measure Yes Yes New items and combined in

behavioural measure. Only observer

data used for model testing.

The expected relationships between the measured variables were discussed in

Chapter 3. It is important to note that some conceptual overlap may occur between the

scales in Table 5.5. For example, the OCB scale developed by Podsakoff et al. (1990),

reports to measure Organ’s (1988) five factors of OCB – altruism, conscientiousness


(alternatively named – compliance), sportsmanship, courtesy, and civic virtue. Altruism

involves helping other people, which is also true for individual’s high in agreeableness.

Additionally, people who are agreeable try to avoid conflict, which is consistent with

the OCB sportsmanship factor. The OCB factor of conscientiousness would be expected

to have conceptual overlap with the personality trait of conscientiousness; while both

would be expected to be related to commitment and job involvement. While these

conceptual overlaps are expected, it is important to recognise that the constructs are

relatively distinct and exist at different levels of the personality hierarchical structure

(Eysenck, 1998), as discussed in Chapter 3. However, it is difficult to measure these

constructs without a certain degree of overlap.

Procedure

A data collection programme was developed through liaison with participating

Ships and Establishments. To achieve this, each nominated Ship and Establishment was

contacted through written correspondence to the Commanding Officer requesting

possible data collection times. Once approved to conduct the data collection, Petty

Officers were notified of times that they were able to volunteer for the study. The data

collection sessions were scheduled in two hour periods, as indicated by the trial.

I travelled to all data collection sites to administer all of the self-report

measures. These measures were administered to 202 Petty Officers during October 2006

through to December 2006, in groups of one to 30 individual/s, depending on the

availability of the Petty Officers. During data collection, signed research Informed

Consent Forms were gained. The SPM and NEO PI-R completion instructions were

read out loud to the participants as per the instructions in the relevant instruction

manuals. This was followed by general completion instructions for the Behavioural


Questionnaire and PANAS. I remained in the assessment venue throughout the

assessment. Once participants had completed the questionnaires, responses were

reviewed for missing data. Participants were requested to complete any missing data

that was identified prior to exiting the assessment venue. If further missing data were

discovered during data entry, the participant was contacted.

Petty Officers were asked to identify a supervisor, peer, and subordinate to

conduct an OCB assessment on the Petty Officer, as part of the 360 degree feedback

process. Petty Officers were given the observer behavioural surveys with research

information and consent forms to give to their nominated supervisor, peer and

subordinate. Petty Officers were requested to ask these personnel if they would like to

volunteer to participate in this study. If so, the observers were to complete the

questionnaire and post the completed forms using the supplied self-addressed

envelopes. Their completed forms were forwarded directly to myself and not returned to

the Petty Officer to ensure data were not adjusted. I received the completed observer

questionnaires between October 2006 and June 2007. Observer responses included 169

supervisors, 159 peers, and 163 subordinates who volunteered to complete the

behavioural report on a participating Petty Officer.


CHAPTER SIX – RESULTS

The main aim of the current study was to develop and test a model to predict

OCB. One of the major advances of the study’s design was to incorporate 360 degree

feedback to collect data from multiple sources for the variables of OCB and

interpersonal skills. This data collection strategy has been increasingly utilised in

behavioural research and more specifically for investigating overall workplace

performance. However, it has been rarely utilised in the design of OCB prediction

studies. As such, prior to the model testing, a preliminary investigation was conducted

to analyse the similarity of rater scores for OCB and interpersonal skills in the 360

degree feedback data collection strategy.

Before any analysis was conducted, data were screened for accuracy of input.

Missing data and accuracy of data entry were assessed using SPSS Frequencies. No

missing score was identified and no score exceeded the possible maximum or minimum

value for each variable. An additional review was conducted to check for the possibility

of data entry error. This involved randomly sampling scores from 30 cases for each

variable. All data in these cases was re-checked and no input errors were detected.

Notably, 15 cases had score discrepancies for more than two of the five sets

from the Standard Ravens Progressive Matrices, which could produce invalid results for

individual psychological assessment (Raven, Raven, & Court, 2000). However, Raven

et al. indicate where the data are being used for general research purposes rather than

individual assessment, the overall score is considered sufficiently valid. Consequently,

all 15 cases were retained for analysis. The final data sample comprised of 147

complete data sets, as previously indicated in Chapter 5.


360 Degree Feedback Rating Source Analysis

The current study provided an opportunity to extend the rating source similarity

research conducted Allen et al. (2000). Specifically, while Allen et al. compared the

similarity of OCB rating responses between self, supervisors, and subordinates, the

current study also includes peer responses. Furthermore, the rating similarity analysis is

replicated for the construct of interpersonal skills. The following sections discuss the

assumption testing and the results of the similarity analysis for OCB and interpersonal

skills ratings.

OCB Rating Source Data Assumption Testing

The assumption testing conducted prior to the analysis for similarity of OCB

rating sources included ensuring that there were no univariate outliers, the variables

were normally distributed, there were linear relationships between variables, and

homoscedasticity was present. One OCB rating source standard score was slightly in

excess of ±3.29 (p<0.001), which is the level indicating the possible presence of a

univariate outlier (Tabachnick & Fidell, 2007). A further review of this case revealed

that the response was made by a subordinate and the other rating scores for this case

were within ±3.29. As such, it was decided that the deletion of this case was not

necessary.

Univariate normality was assessed initially using skewness and kurtosis scores,

which ranged between -0.83 and -0.04 for skewness, and -0.10 and +1.37 for kurtosis.

Notably, all variables were less than the ±1.96 tolerance levels recommended by

Tabachnick and Fidell (2007). However, in a further test to assess normality, the

Kolmogorov-Smirnov statistic was significant at p<0.01 for supervisor and peer


suggesting these scores were non-normally distributed. However, large sample sizes can

produce a significant result from small deviations in normality in the Kolmogorov-

Smirnov statistic (Tabachnick & Fidell, 2007). In such cases, Tabachnick and Fidell

recommend a visual inspection of histograms, which indicated that there was no severe

deviation from normality in the scores for each of the rating sources. Moreover, as the

skewness and kurtosis scores indicated relatively normal distributions for all rating

sources, the assumption of normality was considered sufficiently achieved.

The assumption of linearity was initially assessed using scatterplots of all rating

pairs to rule out gross non-linear relationships. A visual inspection of the scatterplots

indicated that the scatterplots for self-rating and each of the other raters suggested that

the variables conformed adequately to the assumption of linearity and no data

transformations were undertaken.

The final assumption of homogeneity of variance was assessed using Levene’s

test, as recommended by Tabachnick and Fidell (2007). The observed Levene statistic

was not significant (p>0.05), indicating that the variance in each of the rater scores was

sufficiently similar to meet the homogeneity of variance assumption.

Interpersonal Skills Rating Source Data Assumption Testing

The assumption testing that was conducted for the OCB scores was also

conducted for the interpersonal skills ratings. Four interpersonal skills rating source

standard scores were in excess of ±3.29 (p<0.001) indicating the possible presence of a

univariate outlier, based on the criteria suggested by Tabachnick and Fidell (2007).

Inspection of each of these cases identified that one case was for the self-rater (+3.47),

one case for the peer rater (-3.96), and two cases for the subordinate rater (-3.76; -3.54).

However, similar to the OCB rating score that was identified as a possible univariate


outlier, the interpersonal skills scores provided by the other raters for each case were

within ±3.29. As the standard scores were only slightly in excess of ±3.29, once again it

was argued that it was not necessary to delete these cases.

Univariate normality for the interpersonal skills rating scores was assessed

initially using skewness and kurtosis scores, which ranged between -0.96 and 0.13 for

skewness, and 0.13 and 1.86 for kurtosis. Notably, all raters’ scores were less than the

±1.96 tolerance level recommended by Tabachnick and Fidell (2007). However, similar

to the results of OCB rater scores, the Kolmogorov-Smirnov statistic was significant

(p<0.01) for self-rater and subordinate rater; while the supervisor rater and peer rater

were non-significant. These results indicated that self and subordinate rating scores

could be non-normally distributed; while the supervisor and peer ratings met the

assumption of normality. However, as previously discussed, large samples can create a

significant result from small deviations in normality in the Kolmogorov-Smirnov

statistic (Tabachnick & Fidell, 2007). As recommended by Tabachnick and Fidell

(2007), visual inspection of the histograms was conducted, which indicated no severe

deviation from normality in the interpersonal skills scores for each rating source. As the

skewness and kurtosis scores indicated relatively normal distributions for all rating

sources, the assumption of normality was considered sufficiently achieved.

The assumption of linearity was assessed using scatterplots of all rating pairs to

rule out gross non-linear relationships. A visual inspection of the scatterplots indicated

that the scatterplots for all pairs of raters were sufficiently linear. This indicated that the

relationships between the variables conformed adequately to the assumption of linearity

and no data transformations were undertaken.


The final assumption of homogeneity of variance was assessed using Levene’s

test. The observed Levene statistic was significant (p<0.01), indicating that the variance

in each of the rater scores was not sufficiently similar and violated the homogeneity of

variance assumption. The violation of this assumption indicated that a more robust F-

ratio be used to assess the difference in the means of the raters. As such, the one-way

ANOVA F-test statistic was not used for the interpersonal skills rating data, but rather

the data were tested using the Brown-Forsythe F-ratio, as suggested by Field (2009).

The results of the analysis are described in the following section.

Congruence-r Rating Source Analysis

The similarity rating source analysis was based on the work of Allen et al.

(2000). When comparing the OCB and interpersonal skills scores from self, supervisors,

peers, and subordinates, the following hypotheses were assessed:

Hypothesis 1: The strongest correlation for OCB scores will exist between supervisor

and peer ratings.

Hypothesis 2: The strongest correlation for interpersonal skills scores will exist

between supervisor and peer ratings.

To test Hypothesis 1, the current study conducted correlational analysis between

each of the rating pairs. The correlations together with other descriptive statistics for all

OCB rating sources are shown in Table 6.1. As noted previously, the scale to measure

OCB for self-reports consisted of five items. In contrast, the study utilised the Podsakoff

et al. (1990) 24-item OCB scale to measure the responses from the other observers. As

such, a mathematical transformation was applied to the mean self-score to produce a

comparable mean score which was 105.96 (SD=13.60). The reliability estimates for all


scales were high (α≥0.92) except for the self-report measure of OCB that had a lower

reliability estimate (α=0.56); which may be due partially to the use of only five items.

To further assess why the reliability of the self-report scale for OCB was low,

further investigations were conducted on each of the five items. The five items

measured the OCB aspects of helping others, volunteering for non-compulsory

committees and meetings, maintaining a high work effort, courtesy, and supporting

organisational policies even if the individual did not believe in them. Notably, the

aspects of volunteering for non-compulsory committees and meetings (M=17.21,

SD=6.07) and supporting organisational policies even if the individual did not believe in

them (M=19.85, SD=5.11) had high standard deviations, while the other three aspects

were reasonable. It is probable that due to the variance in these two items, the self-

report OCB scale had a reliability score that was lower than expected. However, it is

important to note that the removal of these two items would further reduce the

reliability of the OCB self-report scale. As such, each of the five items were retained,

with an acknowledgement that the two identified OCB aspects are not consistently

responded to within the sample.

Table 6.1 Descriptive Statistics and Correlations for the Organisational Citizenship

Behaviour Observers: Self (5 items), Supervisors (24 items), Peers (24 items), and

Subordinates (24 items) (N=147)

* Correlation is significant at the 0.05 level (1-tailed)

** Correlation is significant at the 0.01 level (1-tailed)

OCB Rating Source Mean SD Alpha 1 2 3

1. Self-Rating 105.96 13.60 .56

2. Supervisor Rating 111.17 14.02 .94 -.01

3. Peer Rating 112.39 11.11 .92 .14* .47**

4. Subordinate Rating 114.52 14.01 .94 .06 .25** .28**


The correlations in Table 6.1 demonstrated that the strongest OCB rating source

correlation was between supervisors and peers (r=0.47, p<0.01); which provided initial

support for Hypothesis 1. To determine if the supervisor-peer relationship was

statistically stronger than the other correlations, I tested for statistical differences

between the correlations, as suggested by Field (2009). Notably, the supervisor-peer

correlation was significantly larger (p<0.05) than the correlations of supervisor-self,

supervisor-subordinate, self-peer, and self-subordinate. However, there was no

significant difference between the supervisor-peer correlation and the peer-subordinate

correlation. As such, partial support was achieved for Hypothesis 1. Specifically, the

strongest source similarity rating occurred between the supervisor and peer, and this

was higher than all the other correlations other than the peer-subordinate correlation.

The strength of the supervisor-peer correlation has been previously observed by Harris

and Schaubroeck (1988), Conway and Huffcutt (1997), Mount et al. (1998b), and Warr

and Bourne (2000). As mentioned, Allen et al. (2002) had observed the highest

correlation between rating sources was the pairing supervisors and subordinates;

however, their study did not include the ratings of peers. Notably, the supervisor-

subordinate correlation was the highest correlation in this study if peer ratings were

removed.

To assess Hypothesis 2, the current study conducted congruence-r analysis

between all rating source pairs of interpersonal skills. Descriptive statistics and

correlations for all OCB rating sources are shown in Table 6.2. The reliability estimates

for all scales were high (α≥0.80).


Table 6.2 Descriptive Statistics and Correlations for the Interpersonal Skills Raters (All

Raters - 14 Items, N=147)

* Correlation is significant at the 0.05 level (1-tailed)


Similar to the results for OCB, the strongest interpersonal skills ratings

correlation was observed between supervisors and peers (r=0.39), initially suggesting

support for Hypothesis 2. However, in contrast to the correlations observed in the OCB

rating analysis, all correlations were significant. Notably, further analysis comparing the

supervisor-peer correlation with the other correlations failed to show any significant

differences. This result suggested that all rating sources rated the Petty Officer similarly.

As such, the current study did not observe support for Hypothesis 2. Notably, this result

is consistent with Conway & Huffcutt (1997) who also observed higher between-rater

agreement for interpersonal skills. Allen et al. (2000) provided an explanation for the

result as interpersonal skills are commonly assessed as part of the annual performance

review for Petty Officers. As all raters have experienced the annual performance review

process, the raters would have a shared understanding of the rating process for this

variable thereby increasing the similarity between rating scores.

OCB Rating Source Mean SD Alpha 1 2 3

1. Self-Rating 61.17 6.58 .80

2. Supervisor Rating 62.47 8.36 .90 .30**

3. Peer Rating 64.06 6.59 .86 .20** .39**

4. Subordinate Rating 65.29 9.12 .92 .29** .36** .30**


Congruence-d Rating Source Analysis

Following the congruence-r analysis, the current study investigated the similarity

of raters’ OCB and interpersonal skills scores by assessing the mean difference between

the raters’ scores. The two hypotheses that were assessed based on this analysis were:

Hypothesis 3: There will be a greater difference between the OCB scores of self and

others than between the other source raters.

Hypothesis 4: There will be a greater difference between the interpersonal skills scores

of self and others than between the other source raters.

Hypothesis 3 was assessed by conducting a one-way ANOVA to compare the

mean difference between ratings sources. The ANOVA analysis was significant

(F(3,584)=11.12, p<0.01), indicating the mean ratings differed significantly across the

raters.

To determine where the difference between the rating sources occurred, a

Tukey’s HSD test was conducted. The results demonstrated that the significant

difference in the means was between self-raters and each of the other raters and not

between the other raters. As such, the results of Tukey’s HSD test supported Hypothesis

3. However, the difference between self and other raters was negative, which indicated

that self-raters underscored themselves on OCB. This finding contrasts with the results

of Harris and Schaubroeck (1988) and Warr and Bourne (2000), who suggested that

self-rating would be higher than the ratings from other observers due to self-interest.

Moreover, Warr and Bourne observed that “easy” behaviours in comparison to technical

behaviours are over-rated by self-raters and under-rated by other observers. As OCB


would be an “easy” behaviour, as classified by Warr and Bourne, self-raters would be

expected to over-rate their performance on OCBs.

The under-rating of OCB by self-raters in this study can possibly be explained

through cultural influences. Specifically, in comparison to the previously reported

research (e.g. Allen et al., 2000 – United States sample; Warr & Bourne, 2000 – United

Kingdom sample), the current study utilised a sample of Royal Australian Navy Petty

Officers. Australians are culturally expected to behave with humility and not see

themselves as better than others (Boyd & Lees, 2012). As such, the Australian culture

could influence the self-raters to under-rate their performance in comparison to the

ratings of others. In addition, the cultural influence of humility could be even stronger

in the Royal Australian Navy sample, particularly when considering issues of teamwork

and mateship. The cultural influences on the current sample are further discussed in

Chapter 7.

The fourth hypothesis suggested there would be a greater difference between the

interpersonal skills scores of self and others than between the other raters. To assess for

differences in the mean of rating sources as previously discussed, the current study

utilised the Brown-Forsythe F-ratio, which is robust when the homogeneity of variance

assumption was violated (Field, 2009). The results of the Brown-Forsythe F-ratio was

significant (F(3,539)=7.96, p<0.01), indicating the mean ratings differed significantly

across the raters. This result is consistent with the findings observed by Allen et al.

(2000).

To determine where the difference occurred, a Tukey’s HSD test was conducted.

The results indicated partial support for Hypothesis 4. Notably, differences in the means

occurred between self and peers and self and subordinates, however there were no


significant differences in the means of self and supervisors. Once again, it is argued that

this result is due to self and supervisors possessing similar expectations of interpersonal

skills due to the past experience of annual performance reviews (Allen et al., 2000).

Consistent with the results observed in the OCB analysis, the difference between self

and both peers and subordinates produced an underscoring of interpersonal skills by the

self-rater. Once again this result is explained through cultural influences, which as

further discussed in Chapter 7.

Summary of Rating Source Analysis for OCB and Interpersonal Skills

The current study assessed four hypotheses relating to the 360 degree feedback

rating source scores for the constructs of OCB and interpersonal skills. The hypotheses

investigated the level of similarity in rater responses, both with respect to congruence-r

and difference of means. Results indicated that the strongest relationships between OCB

raters were the supervisor-peer and the peer-subordinate correlations. This result

provided partial support for Hypothesis 1. In contrast, no support was observed for

Hypothesis 2, as all correlations were statistically similar for interpersonal skills ratings.

The rating source similarity analysis observed support for Hypothesis 3 and partial

support for Hypothesis 4. Specifically, for OCB scores results indicated that the mean

for self-raters was different from the mean of other raters and there was no difference

between the other raters of OCB; while a difference between interpersonal skills score

means was observed for self-raters with peers and subordinates, albeit not for supervisor

ratings. The results of Hypothesis 4 are explained due to interpersonal skills being

included in annual performance reviews, which may have increased the similarity of

behavioural expectations between the Petty Officers and the supervisor (Allen et al.,

2000).


The pattern of results across the four hypotheses suggested that OCB and

interpersonal skills were considered different variables by each of the raters.

Interestingly, the direction of difference for self-raters for both OCB and interpersonal

skills was under-rated by the Australian Navy population in comparison to the other

raters. The direction of appraisal is in contrast to Harris and Schaubroeck (1988) and

Warr and Bourne (2000). As mentioned, this result and its implications will be

discussed in more detail in Chapter 7.

Structural Equation Modeling

Structural Equation Modeling (SEM) is a “collection of statistical techniques

that allow examination of a set of relationships between one or more IVs [independent

variable], either continuous or discrete, and one or more DVs [dependent variable],

either continuous or discrete” (Tabachnick & Fidell, 1996, p.709). SEM is a statistical

approach to conduct confirmatory analysis, enabling the researcher to simultaneously

investigate multiple theoretically-based relationships between variables. Byrne (2010)

argues that these relationships can provide a clearer theoretical picture through the use

of diagrammatic representation. It is a powerful multivariate approach that combines

path analysis and factor analysis. As a model usually consists of multiple variables and

indicators, SEM is more complicated than either of these two statistical methods alone.

The purpose of SEM is to test the accuracy of a hypothesised model against a

sample population. From a mathematical perspective the data is equal to the model plus

the residual, where the residual is the difference between the observed data and the

hypothesised model (Byrne, 2010). The smaller the residual, the more representative the

model is of the observed data. The model is statistically assessed through numerous

goodness-of-fit indices. Many goodness-of-fit indices have been developed and the


researcher should aim to select a range of indices that would be appropriate to assess the

fit or otherwise of the hypothesised model. The indices selected for the current research

will be discussed in the Model Testing section of this chapter.

Assumption Testing for SEM

Assumption testing for SEM initially involved assessing the variable distribution

and also checking for the possibility of univariate outliers. Univariate normality was

sufficiently achieved as all variables had skewness and kurtosis scores that were less

than the ±1.96 tolerance levels, as recommended by Tabachnick and Fidell (2007). All

variables, with the exception of conscientiousness, had standard scores less than 3.29

(p<0.001), which as previously discussed in the rater-source similarity analysis of this

chapter, indicated the possibility of univariate outliers. Conscientiousness had one case

with a standard score (3.35) in excess of 3.29. Notably, this case did not have scores in-

excess of 3.29 for the other scales. As the case was only slightly in-excess of the

tolerance level, it was decided the case would be retained.

SEM techniques assess linear relationships between model variables

(Tabachnick & Fidell, 2007). The assumption of linearity was initially assessed using

scatterplots of all rating pairs to rule out gross non-linear relationships. A visual

inspection of the scatterplots indicated relatively linear relationships between the paired

variables, with the exception of the pairing of agreeableness and interpersonal skills,

which produced a dispersed pattern of points. Given that the scatterplots did not suggest

gross non-linear relationships, no data transformations were undertaken. These

relationships were further assessed using correlational analysis. The results of this

analysis, in addition to the descriptive statistics for the hypothesised model variables,

are produced at Table 6.3.


Varia

ble

M

ean

S

D

Alp

ha

1

2

3

4

5

6

7

8

9

1

Gen

eral

Men

tal

A

bil

ity (

Self

)

40

.63

23

.21

.83

- .

93#

2

OC

B

(

Ob

serv

ers)

33

8.0

9

29

.14

.95

.15

3

In

terp

erso

nal

S

kil

ls (

Ob

serv

ers)

19

1.8

2

18

.05

.92

.17

*

.81

**

4

Agree

ab

len

ess

(

Self

)

41

.30

9.4

9

.72

-.1

8*

-.0

7

-.1

0

5

Con

scie

nti

ou

sness

(

Self

)

42

.82

10

.13

.86

-.0

1

.26

**

.2

3**

-.

01

6

Extr

aversi

on

(

Self

)

43

.97

10

.83

.81

-.1

3

.08

.12

-.0

2

-.1

1

7

Posi

tive A

ffect

(

Self

)

42

.37

6.2

0

.88

-.0

7

.17

*

.15

-.0

5

.39

**

.4

4**

8 Job

Sati

sfacti

on

(

Self

)

99

.29

10

.21

.79

-.0

5

.08

.08

.11

.18

*

.11

.44

**

9

Job

Sati

sfacti

on

W

eig

hte

d (

Self

)

43

7.0

5

71

.16

.76

-.0

3

.12

.15

.08

.29

**

.2

1**

.5

1**

.8

5**

10 Job

In

volv

em

en

t

(

Self

)

78

.10

10

.39

.79

-.1

1

.13

.01

.02

.45

**

.0

7

.45

**

.3

8**

.4

5**


Means, standard deviations, reliability estimates, and correlations for all model

variables are shown in Table 6.3. Reliability estimates achieved for this study across all

scales were adequate (α≥0.72). The observers rating of OCB, which aggregated the

scores from supervisors, peers, and subordinates, gained the highest reliability estimate

(α=0.95). High reliability estimates were also achieved for the observer rating of

interpersonal skills (α=0.92). The reliability estimates for job involvement and job

satisfaction scales are presented in Table 6.3. Notably, the reliability for job satisfaction

corresponds to those previously demonstrated by Cook et al. (1981). The new scale of

weighted job satisfaction also achieved adequate reliability estimates (α=0.76). The

reliability estimates for extraversion (α=0.81) and conscientiousness (α=0.86) were

good, while agreeableness had the lowest reliability estimate of variables included in the

hypothesised model (α=0.72). An alpha of 0.88 was produced for the positive affect

scale in the PANAS, which was consistent with the reliability reported in past studies

that have utilised the PANAS (Watson et al., 1988).

The correlations between the variables that are described at Table 6.3 provide an

opportunity to review the linear relationships to be tested in the model. The

hypothesised model, at Figure 3.12, theorised relationships between agreeableness and

interpersonal skills and also extraversion and interpersonal skills. Notably, as can be

seen in Table 6.3, neither of these relationships was significant for this sample.

However, conscientiousness and general mental ability were observed to have small but

significant correlations with interpersonal skills (r=0.23; r=0.17, respectively).

Extraversion and conscientiousness were observed to have significant correlations with

positive affect (r=0.44; r=0.39, respectively). As hypothesised, conscientiousness also

had positive significant relationships with job satisfaction and job involvement (r=0.18;

r=0.45, respectively). Furthermore, positive affect had positive relationships with job


satisfaction and job involvement (r=0.44; r=0.45, respectively). Interpersonal skills had

a very strong positive relationship with OCB (r=0.81), while the relationship between

job involvement and OCB was non-significant. As observed in Table 6.3, job

involvement had the expected significant positive relationships with conscientiousness

(r=0.45), positive affect (r=0.45), and job satisfaction (r=0.38).

Given that agreeableness and extraversion did not have the hypothesised

relationships with interpersonal skills, further investigations were conducted.

Specifically, the six facets of agreeableness were assessed to determine if any correlated

with the overall score of interpersonal skills, but none did. The absence of significant

relationships supports the perspective that agreeableness is not an important predictor of

interpersonal skills, as previously reported by Ferris et el. (2001). As such, the

agreeableness variable was removed from the hypothesised model.

Correlational analysis was completed for the six extraversion facets with the

variables of interpersonal skills and positive affect. As shown in Table 6.4, the

extraversion facets of E3 – assertiveness and E4 – activity were observed to have small

but significant positive relationships with interpersonal skills (r=0.25; r=0.23,

respectively). Moreover, all of the extraversion facets had significant relationships with

positive affect. These results suggested that extraversion, or a part of it, was related to

interpersonal skills and more strongly with positive affect. As such, it was decided that

the complete extraversion score would be retained in the hypothesised model.


Table 6.4

Extraversion Facet Correlations with Interpersonal Skills and Positive Affect

Extraversion Facet Correlation with


Correlation with

Positive Affect

1 Warmth .10 .43**

2 Gregariousness .05 .35**

3 Assertiveness .25** .48**

4 Activity .23** .46**

5 Excitement Seeking .06 .24**

6 Positive Emotions .11 .40**


In addition to assessing the linear relationship between the hypothesised model’s

variables, the correlations shown in Table 6.3 provided the opportunity to assess the

multicollinearity and singularity assumption required for SEM. According to

Tabachnick and Fidell (1996), multicollinearity occurs when the independent variables

achieve correlation values greater than 0.90; while singularity is present if the

correlation is 1.0. They argue that high to perfect correlations creates redundancy in the

analysis as the two variables are essentially measuring the same information. Assessing

multicollinearity and singularity have benefits for model parsimony, which refers to

reducing the complexity of the model through a reduction of variables, while retaining

appropriate goodness-of-fit measures (Tabachnick & Fidell, 1996). While no correlation

between independent variables were greater than 0.90 suggesting that no

multicollinearity or singularity was present, the correlation between job satisfaction and

weighted job satisfaction was high (r=0.85). Moreover, both variables correlated

similarly with the other variables in the model. The relationship between job satisfaction

and weighted job satisfaction was expected to be strong as both variables were based on

the original job satisfaction score. Thus only one form a job satisfaction needed to be


included in the OCB prediction model. As similar scores were observed for both

variables, the established measure of job satisfaction was selected for model analysis.

The remaining independent variables were assessed for multicollinearity using

OCB as the dependent variable and evaluating the two collinearity diagnostic factors of

tolerance and variance inflation factor (VIF). The independent variables assessed

included interpersonal skills, general mental ability, extraversion, conscientiousness,

positive affect, job satisfaction, and job involvement. Tolerance values ranged between

0.49 and 0.94, which were greater than 0.1 suggested by Kline (2011) as the minimum

level that would indicate multicollinearity. Moreover, VIF ranged between 1.07 and

2.06, which was lower than the value of 10, as recommended by Kline, which also

indicated no multicollinearity between the independent variables.

It is important to note that the observer ratings for OCB and interpersonal skills

strongly correlated (r=0.81). As the relationship was between an independent variable

(interpersonal skills) and the dependent variable (OCB), multicollinearity was not

relevant (Tabachnick & Fidell, 2007). However, it was necessary to determine if both

scales were measuring the same or different constructs. Organ et al. (2006) suggest that

the Podsakoff et al. (1990) scale is commonly used, providing evidence that the scale

measures OCB. In contrast, the interpersonal skills scale was newly developed for the

purpose of this study and was based on a scale developed by Ferris et al. (2001). Ferris

et al. observed a correlation of 0.20 between interpersonal skill and OCB, albeit

alternatively named in their study, suggesting that they had measured two constructs

that held a small correlation. Further evidence of the measurement of two constructs

was obtained in the rating source analysis section of the current chapter. Specifically,

interpersonal skills and OCB measures obtained different rating patterns by the rating

sources. However, the current study would suggest that a great deal of similarity was


occurring in the measurement of these two constructs. Given that past studies (e.g.

Ferris et al., 2001) have indicated that OCB and interpersonal skills are different

constructs, it was decided to retain both constructs in the model analysis, despite the

stronger than expected correlation. The practical significance of the existence of the

very strong correlation will be discussed further in Chapter 7.

Given the findings of the correlational analysis, the final variables to be tested in

the hypothesised model were OCB, interpersonal skills, general mental ability,

conscientiousness, extraversion, job involvement, job satisfaction, and positive affect.

The modified hypothesised model from Figure 3.12 is presented at Figure 6.1. The

variables in this model underwent the final two assumption tests of multivariate

normality and adequacy of sample size. Byrne (2010) argues that multivariate normality

is a critically important assumption of SEM and in particular the presence of

multivariate kurtosis is of concern to SEM researchers. To assess the data for possible

multivariate non-normality, two methods were employed. Firstly, the Mahalanobis

Distance was used to identify possible multivariate outliers. This type of outlier would

have statistically extreme scores on two or more variables (Kline, 2005). Tabachnick

and Fidell (1996) recommend the conservative probability estimate value of less than

0.001 to identify possible outliers. No multivariate outliers were detected, as the

assessed squared Mahalanobis Distance values were found to have probability scores

greater than 0.001. Secondly, the critical value (CR), which should be less than five,

according to Bentler (2005), was assessed using AMOS. The CR was 3.49 for the

hypothesised model shown in Figure 6.1. Considering these findings from these two

assessments, the assumption of multivariate normality was considered to be satisfied.


Job

Satisfaction

Extraversion

Conscientiousness

Positive

Affect



OCB

Job Involvement

E1

E2

E3

E4E5

E7

E81

1

1

1

1

E6

1

1

1

The final assumption for SEM is that there is an adequate sample size, as an

insufficient sample size can decrease the reliability of the SEM analysis (Ullman, 1996).

The data sample to assess the final hypothesised model, as shown at Figure 6.1, retained

the same 147 cases that were utilised in the 360 degree analysis. The model has eight

observed variables and 20 estimated parameters producing ratios of 18:1 for participants

to observed variables and 7:1 for participants to estimated parameters. Ullman (1996)

considers that these ratios are sufficient to conduct SEM and therefore the data sample

of 147 is adequate.

Model Testing

Following the data cleaning, screening and assumption testing, the data were

considered acceptable to conduct SEM. Jöreskog (1993) indicates there are three main

approaches to conducting SEM. The first is Strictly Confirmatory, where a single

Figure 6.1.

Final hypothesised model with error terms: Predicting overall job performance through

interpersonal skills and job involvement pathways.


hypothesised model is tested and either accepted or rejected based on the goodness-of-

fit indices. There are no modifications if the model is rejected. The second approach is

Alternative Models, which involves the testing of more than one model and a

comparison is made between various models to determine the best fitting model. The

final approach is known as Model Development, and involves post hoc modifications to

a hypothesised model when it is rejected, or to improve an accepted model. I adopted

the first SEM approach of Strictly Confirmatory in the current study. Byrne (2010)

makes the critical point that SEM only tests if the model possessed or lacks the

statistical characteristics to fit the observed data. SEM does not assess if the model is

theoretically sound – that role belongs solely with the researcher. The development of

the hypothesised model requires theoretical reasoning.

Goodness-of-fit indices assess if the model is statistically accepted or rejected.

There are numerous goodness-of-fit indices that could be used in the analysis process

and each of the indices provides a different statistical perspective. Two important forms

of indices are absolute fit indices and parsimony fit indices. Byrne (2010) states that the

absolute fit indices compare the hypothesised model with no model. In contrast to

traditional statistical procedures, the null hypothesis being tested is also the

hypothesised model. In SEM for absolute fit indices, the researcher hopes not to reject

the null hypothesis. In contrast the parsimony fit indices consider the complexity of the

model, preferring the simple over the complex. Five goodness-of-fit indices were

chosen to provide the necessary variation in statistical perspectives to determine if the

hypothesised model was accepted or rejected. The selected absolute fit indices were the

minimum discrepancy (CMIN), Goodness-of-Fit Index (GFI), Adjusted Goodness-of-

Fit Index (GFI), and Comparative Fit Index (CFI). The Mean Square Error of

Approximation (RMSEA) was selected as the parsimony fit index. These statistics will


now be described; including the minimum values that would be required for the

hypothesised model to be indicative of a good fit with the observed data.

The original absolute fit index is the CMIN and represents the Likelihood Ratio

Test statistic, more commonly known as Chi Square (2) (Byrne, 2010). In contrast to

traditional statistics, a statistically significant value indicates that the data is not

representative of the model. Conversely, if the 2

value is associated with a probability

of greater than 0.05, the hypothesised model is supported. It is noted that while CMIN is

the most commonly used index; it can be affected by sample size, model size,

distribution of variables, and omitted variables. As such, it is generally only used as a

quick initial assessment of the model and other goodness-of-fit indices are required to

assess the model fit.

Two additional absolute goodness-of-fit indices are the GFI and AGFI, which

are statistical variations to the 2 statistic. These additional indices enable the researcher

to consider the acceptance or rejection of the model in conjunction with the CMIN. The

GFI is a “measure of the relative amount of variance and covariance in S [variance-

covariance matrix of the sample data] that is jointly explained by [variance-

covariance matrix for the hypothesised model]” (Byrne, 2010, p.77). Simply, the GFI is

the degree of fit of the squared residuals from the hypothesised model as compared to

the actual data that was collected. The AGFI adjusts the GFI by compensating for the

degrees of freedom of the model. AGFI therefore also considers the parsimony of the

hypothesised model. Scores for both indices can range between 0 and 1, with 1

indicating a perfect fit (Arbuckle & Wothke, 1999). However, as both the GFI and the

AGFI are statistical variations of the 2 statistic, they can also be affected by sample

size. As such, additional indices are required.


SEM is generally a large sample size analysis technique and as such some

goodness-of-fit indices, such as the GFI and AGFI, can underestimate the fit of the

model in small sample sizes. The current study has a smaller sample size and requires a

goodness-of-fit index that provides increased sensitivity for small sample sizes.

Consequently, the chosen index was the CFI. The CFI is also an absolute goodness-of-

fit index, which takes into account small data sample sizes. This index is derived

through a comparison between the hypothesised model and the null hypothesis. Values

range between 0 and 1, with 1 indicating a perfect fit for the hypothesised model. Hu

and Bentler (1999) indicate that a score greater than 0.95 is required to demonstrate a

good fitting model.

The final goodness-of-fit index that will be used in this study is the Root Mean

Square Error of Approximation (RMSEA), which considers the parsimony of the model.

Byrne (2010) suggests this index is recognised as one of the most informative amongst

the goodness-of-fit indices as it considers the error of approximation in the population.

For the hypothesised model to be indicative of a good fit, RMSEA must be less than 0.1

and preferably less than 0.05 (Browne & Cudeck, 1993).

The tested model shown in Figure 6.2 graphically represents the complex

interactions of casual pathways of the individual variables to predict OCB. The model

includes the standardised path coefficients, which represent the standardised total

effects between variables (Kline, 2011). The variables of general mental ability,

conscientiousness, extraversion, positive affect, job satisfaction, and job involvement

were measured using self-report data from the Petty Officers. Three observer ratings

were gained from a supervisor, a peer, and a subordinate for each Petty Officer to

measure the variables of interpersonal skills and OCB. An overall observer score was

created by aggregating each of the three observers for each case for these two variables.


The number of observed variables was eight, with estimated parameters totalling 20.

SEM was used to assess this model using AMOS 17.0.0 (Arbuckle, 2008). The 2

statistic supported the hypothesised model (2=12.19, 16, p>0.05). Inspection of the

remaining indices also suggested acceptable fit (GFI=0.98; AGFI=0.96; CFI=1.00;

RMSEA=0.00). Model analysis supported the hypothesised model and post hoc testing

was not required.

The OCB prediction model presented at Figure 6.2 accounts for a large amount

of variance in OCB (68%). All standardised regression coefficients in the final model

were positive. The strong relationship between interpersonal skills and OCB (r=0.81)

resulted in the largest standardised regression coefficient occurring for the interpersonal

skills-OCB path (0.80). Notably, the job involvement path to predict OCB was much

smaller (0.12). These results suggest that interpersonal skills are more influential than

job involvement in the prediction of OCB. However, reasonable amounts of variance

.20

Job Satisfaction

.00

Extraversion

.00

Conscientiousness

.42

Positive Affect

.13


.00


.68

OCB

.80

.17

.44

.34

Job Involvement .12

.25

.32

.45

.23

.19

.23

.01

.48

Figure 6.2

Organisational citizenship behaviour prediction model with standardised estimates.

Note: All regression weights within this model were significant (p<0.05), with the exception

of the conscientious – job satisfaction pathway (CR=0.08, p>0.05).


were explained in the variables of job involvement (34%), positive affect (42%), and

job satisfaction (20%). Only a small amount of the variance in interpersonal skills was

explained (13%) indicating that other variables play a role in the prediction of

interpersonal skills.

A review of the regression weights in Figure 6.2 indicated that all pathways

were significant, with the exception of the conscientiousness – job satisfaction pathway

(CR=0.08, p>0.05). Notably, the correlation between these two variables shown in

Table 6.3, had demonstrated a small but significant positive relationship (r=0.18,

p<0.05). The results of the SEM show that the impact of conscientiousness on job

satisfaction has been totally mediated by positive affect (Baron & Kenny, 1986). This

finding suggests that while a direct relationship between conscientiousness and job

satisfaction had been previously theorised (e.g. Organ & Ryan, 1995), the relationship is

actually indirect and mediated by positive affect.

SEM provides an opportunity to consider all of the direct and indirect effects of

the variables on OCB. The most commonly used method to consider indirect effects

within the model was developed by Baron and Kenny (1986). However, increasingly

researchers are utilising statistical programmes to undertake this analysis (Wallace,

Johnson, Mathe, & Paul, 2011). AMOS 17.0.0 provides bootstrapping analysis that can

be utilised to test direct and indirect effects by the model variables (Arbuckle, 2008).

Bootstrapping treats the sample data as a population and involves repeated sampling

from the original data (Blunch, 2008).

A visual inspection of the final model suggested that mediation effects appeared

to be present in both the interpersonal skills and job involvement pathways. The current

study used 1000 bootstrap samples to examine the mediation effects within the model.


The resultant AMOS 17.0.0 output was examined to assess for standardised total, direct,

and indirect effects. Consistent with past studies (e.g. Barrick & Mount, 1991),

conscientiousness was a main contributor to predicting OCB and provided an overall

standardised total effect of 0.26, with standardised direct effects on interpersonal skills

(0.25), positive affect (0.44), job satisfaction (0.01), and job involvement (0.32). As

such, the main contribution of conscientiousness to predicting OCB was through the

positive affect pathway. Extraversion also contributed indirect effects on OCB (0.16),

directly through positive affect (0.48) and interpersonal skills (0.18) in addition to

indirect effects on job satisfaction (0.21) and job involvement (0.16). General metal

ability had a standardised indirect effect on OCB (0.15), which was mediated by

interpersonal skills; while job satisfaction mediated the relationship between positive

effect and job involvement (0.10). The implication of these results will be discussed in

the following chapter.


CHAPTER SEVEN – DISCUSSION

The present study continues the research effort to predict OCB from individual

differences. The emphasis of OCB predictive studies initially focused on relatively

simple relationships between OCB and the variables of either job satisfaction or

personality (e.g. Bateman & Organ, 1983, Organ & Ryan, 1995; Smith et al., 1983).

Further studies assessing the existence of these simple relationships have observed

inconsistent results, as discussed in Chapter 3. One possible reason for these results is

that sufficient research attention has yet to be given to the simultaneous analysis of

multiple variables to predict OCB. This argument is based on the premise that work

behaviour is a complex construct and is determined through a combination of both

personal and situational factors.

The aim of the current study was to identify the additional variables that should

be included in an OCB prediction model, in addition to the traditional antecedents of job

satisfaction and personality. The OCB prediction model was developed based on the

framework provided by Campbell’s (1990) performance model. The selected

antecedents were included in a model that consisted of two major pathways to OCB.

The first pathway was from ability, as measured by interpersonal skills. Interpersonal

skills were identified as the most important ability to predict OCB as this type of ability

involves relationships between people and consists of skills in communication, leading,

teaming, and problem solving (Hayes, 2002; Robbins & Hunsaker, 2006). Motivational

aspects formed the second pathway through job involvement and considered why an

individual would perform OCB within the workplace. The framework for the prediction

of OCB hypothesised that OCB was not directly caused by individual traits but rather


their impact would be mediated by the variables of interpersonal skills or job

involvement.

The variables of job satisfaction, personality traits (conscientiousness,

extraversion, and agreeableness), positive affect, and general mental ability were

included in the interpersonal skills–job involvement framework to predict OCB based

on a collection of correlational and causal findings from previous research, as discussed

in Chapter 3. The design of the study included self-report scales for job involvement,

job satisfaction, personality, positive affect, and general mental ability; while a 360

degree feedback strategy was used for OCB and interpersonal skills. Organ et al. (2006)

recommend a multi-source data collection strategy for observable variables to address

the issue of common method variance. As 360 degree feedback is a relatively new

approach in OCB research, the current study provided an opportunity to observe the

similarity of source ratings and compare the results with the more established use of 360

degree feedback – personnel development (Hedge et al., 2001).

The current study assessed four hypotheses relating to the 360 degree feedback

rating source scores. The hypotheses investigated the level of similarity in rater

responses for OCB and interpersonal skills, both with respect to congruence-r and

difference of means. Results of congruence-r analysis indicated that the strongest rating

similarity was between supervisors and peers. Notably, Allen et al, (2000) suggested

that supervisor and subordinate ratings were the most similar in OCB research;

however, their study did not measure peer ratings. As such, the current research

extended their research and observed results consistent to those produced in personnel

development research (e.g Harris & Schaubroeck, 1988; Warr & Bourne, 2000).


Congruence-d analysis indicated that the OCB mean for self-raters was different

from the mean of all other raters of OCB. In contrast, interpersonal skills rater source

analysis observed that subordinate raters were significantly different to all other raters;

while the self-rated scores were different to the ratings of peers and subordinates but not

with supervisors. The difference of rating response patterns for OCB and interpersonal

skills suggested that the raters considered these two variables as different constructs.

A further finding of interest was the observed under-rating by self-raters in

comparison to the ratings of other sources for both OCB and interpersonal skills. The

direction of appraisal is in contrast to Harris and Schaubroeck (1988) and Warr and

Bourne (2000). This result is explained due to cultural considerations of the Australian

Navy population. National culture has been shown to influence organisational culture,

which affects the employees’ attitudes, decisions, and behaviour in the workplace

(Baird, Harrison, & Reeve, 2007; Sarros, Gray, Densten, & Cooper, 2005). Triandis

(1995) suggests that the Australian culture is horizontally individualistic – individuals

are equal from an egalitarian perspective; while also being independent from each other,

self-reliant, and self-directed. Due to the strong egalitarian beliefs, Triandis suggests

that being different from each other is not valued within the Australian culture.

Moreover, Australian culture expects individuals to behave with humility and not

consider themselves better than their peers (Boyd & Lees, 2012). As such, while self-

report inflation bias was hypothesised (see Harris & Schaubroeck, 1988; Warr &

Bourne, 2000), the Australian self-raters in the current study had a tendency to under-

rate themselves in comparison to the ratings from their supervisors, peers, and

subordinates. This finding provides an interesting direction for future research to

consider if the results are also replicated in other Australian populations, such as private

industry. Notably, Baird et al. observed that similar culture was observed in Australian


service and manufacturing industries. Their observation suggests that national culture

may transcend industry, which can influence the attitudes, decisions, and behaviour in

the workplace. This may include the under-rating of self-behaviour to conform to

cultural norms.

Prior to testing the model, a review of the correlations between the hypothesised

variables in Figure 3.12 was completed. Notably, this analysis suggested that for the

current sample, the personality variable of agreeableness did not contribute to the OCB

prediction model. Agreeableness was included in the hypothesised model as a trait that

would be useful in the prediction of interpersonal skills. Agreeableness describes

individuals who are trustworthy, straightforward, altruistic, complying, modest, and

being tender minded (Costa & McCrae, 1992a). Moreover, as discussed in Chapter 3,

agreeableness has demonstrated relationships with interpersonal skills, especially in

team situations (e.g. Halfill et al., 2008; Mount et al., 1998a; Neuman et al., 1999;

Neuman & Wright, 1999; Tett et al., 1991, and Witt et al., 2002). However, despite the

theoretical reasoning for the inclusion of agreeableness, the theorised relationship

between agreeableness and interpersonal skills was not observed. This result occurred

for both trait and facet levels of agreeableness.

A possible reason for the non-significant result between agreeableness and

interpersonal skills may have been related to the interpersonal skills items. As discussed

in Chapter 4, the interpersonal skills items consisted of a 14 items Likert-type scale.

Five of the 14 items that were used in the current study originated from a seven-item

Likert-type interpersonal skills scale that was developed by Ferris et al. (2001). When

evaluating the seven items used by Ferris and colleagues to measure interpersonal skills

in the current study, I selected five items that were consistent with my construct

definition of interpersonal skills. Moreover, additional items were required to measure


all of the aspects of interpersonal skills that were considered important for the

production of OCB. Specifically, as the five items from the Ferris and colleagues scale

did not measure aspects of conflict resolution or reduction, negotiation ability, helping

others, listening skills, or written communication skills, additional items were included

in the interpersonal skills scale for the current study. As discussed in Chapter 3, while

Ferris and colleagues did not demonstrate a relationship between agreeableness and

interpersonal skills as measured by their seven-item scale; it was hypothesised in the

current study that with the inclusion of the additional items for an interpersonal skills

scale would produce a significant positive relationship. The reasoning for the hypothesis

was that the inclusion of the additional items, particularly the items relating to conflict

avoidance and helping others, have previously demonstrated a relationship with

agreeableness (Jensen-Campbell & Graziano, 2001).

Despite increasing the number of items to measure interpersonal skills including

those that have been shown to be related to agreeableness, the current study did not

demonstrate a significant relationship between agreeableness and interpersonal skills.

Additional correlational analysis between agreeableness facets and interpersonal skills

also failed to produce significant relationships. In reviewing the results of the

hypothesised model analysis, it is concluded that the newly developed interpersonal

skills scale focused more heavily on aspects of communication and sociability rather

than on any of the agreeableness characteristics of empathy, trustworthiness,

straightforwardness, or modesty.

The non-observation of a relationship between agreeableness and interpersonal

skills, as defined in this study, suggested that interpersonal skills were not related to the

trait score of agreeableness for this study. As such, agreeableness was removed from the

final hypothesised model (Figure 6.1) prior to model testing. In contrast, the extraverted


aspects of assertiveness (r=0.25) and activity (r=0.23) were shown to be related to

interpersonal skills. Notably, Chapter 3 discussed that previous research (e.g. Neuman

et al., 1999) has observed a relationship between extraversion and interpersonal skills.

Extraverted people are described as talkative, assertive, active, energetic, outgoing,

outspoken, and dominate (John, 1990). Moreover, as hypothesised, interpersonal skills

were observed to have positive relationships with general mental ability (r=0.17) and

conscientiousness (r=0.23). As such, extraversion, conscientiousness, and general metal

ability were retained in the final hypothesised model (Figure 6.1) in the interpersonal

skills pathway to predict OCB. Correlational analysis for the variables in the job

involvement pathway demonstrated the hypothesised relationships, as presented at

Table 6.3. As such, this section of the model remained unchanged for the model testing

(Figure 6.1).

The hypothesised model was assessed using SEM, which supported the

hypothesised model, with acceptable goodness of fit indices (GFI=0.98; AGFI=0.96;

CFI=1.00; RMSEA=0.00). An interesting finding of the current study was the very

strong bivariate correlation of 0.81 between interpersonal skills and OCB. The strength

of the correlation suggested that the scales for these two constructs were measuring

similar aspects for this sample. Notably, the correlation was much stronger than Ferris

et al. (2001) had observed (r=0.20) in their earlier study. The unexpected result

provides a direction for future research to explore the similarity and differences of the

constructs of OCB and interpersonal skills.

Given the strong bivariate relationship between interpersonal skills and OCB, it

was not surprising that the interpersonal skills pathway (pathway coefficient of 0.80)

was more influential than the job involvement pathway (pathway coefficient of 0.12).

Bootstrapping analysis showed that general mental ability, extraversion, and


conscientiousness indirectly predicted OCB through a mediation effect by interpersonal

skills.

Past research has consistently found that general mental ability provides the best

prediction for overall job performance (Dawis, 2001). The observance of the general

mental ability-interpersonal skills relationship is important, as it is generally observed

when the performance measure is a form of task performance rather than OCB.

Specifically, model analysis from the current study suggests that general mental ability

could also be a small yet valid predictor of OCB; however, interpersonal skills are

required to mediate the relationship.

Although the interpersonal skills pathway to predict OCB was very influential,

the job involvement pathway also demonstrated a small contribution to the prediction of

OCB. The job involvement pathway was formed through a combination of the

variables: job satisfaction, positive affect, conscientiousness, and extraversion.

Consistent with the previously reported relationships of extraversion-positive affect

(e.g. Costa & McCrae, 1980; Lucas et al., 2008; Nemanick & Munz, 1997; and Watson

et al., 1992) and conscientiousness-positive affect (Watson & Clark, 1992b),

extraversion and conscientiousness contributed 42 percent of the variation in positive

affect. While extraversion did not have a significant bivariate correlation with OCB, the

results from the current study demonstrated an indirect role to predict OCB through the

mediating variables of interpersonal skills and positive affect.

The relationship between conscientiousness and OCB (r=0.26) was explained in

the job involvement pathway by the mediation effect of positive affect. A worker high

in conscientiousness would be viewed as efficient, dependable, hard working, and

organised (Costa & McCrae, 1992a). George and Jones (1996) suggested that an


individual with high positive affect maintains higher levels of self-efficacy, possesses

greater expectations of positive outcomes, and is more action-orientated compared to

personnel retaining negative moods. As such, when a conscientious individual

maintains a positive mood, they are more likely to produce OCB. In addition to the

prediction of positive affect, model testing demonstrated that conscientiousness

contributed an indirect contribution to predict OCB through job involvement.

Positive affect contributed indirectly to the prediction of OCB through job

involvement. The current results suggested that high levels of positive affect can

directly influence job involvement, in addition to having an indirect relationship that is

partially mediated by job satisfaction. The partial mediation role of job satisfaction

demonstrates the importance of situational aspects in explaining the positive affect-job

involvement relationship in the prediction of OCB in the workplace. While an

individual may have directly caused job involvement in the workplace through a

positive mood trait, there is also some impact on job involvement based on the

individual’s level of job satisfaction.

The observed partial mediation of the personality-job involvement relationship

by both positive affect and job satisfaction provides further evidence linking personality

and motivational variables when predicting OCB. This is particularly relevant given that

the relationship has been difficult to be consistently obtained, as noted by Gellatly

(1996) that “attempts to empirically link personality characteristics with motivational

variables have produced inconsistent results” (p.474). Specifically, the job involvement

section of the model demonstrated that personality factors (extraversion and

conscientiousness) and affective disposition (positive affect) assist in predicting job

involvement through direct and indirect processes, influencing the production of OCB.


The final model (Figure 6.2) suggested that job involvement was the mediator in

the prediction of OCB for the variables of personality, positive affect, and job

satisfaction. As such, the current study suggested that personality and attitudinal

variables have an indirect role in the prediction of OCB. This finding is consistent with

Organ et al. (2006) who suggested that individual differences, such as personality and

attitudes are more likely to be utilised by the individual in situations where there is

limited direction to the individual on how to behave. Moreover, attitudes and

personality are better at predicting patterns and trends of behaviours rather than specific

behaviour at any given time (Epstein, 1983). Organ et al. (2006) contend that the

prediction of OCB should be considered in terms of attitudes and personality.

In summary, the final OCB prediction model (Figure 6.2) provided an accurate

fit of the data. OCB was predicted through the traditional variables of job satisfaction

and personality (conscientiousness and extraversion) in addition to the variables of

positive affect, general mental ability, interpersonal skills, and job involvement.

Notably, extraversion and conscientiousness contributed to both the interpersonal skills

and job involvement pathways. This result is consistent with Organ and Ryan (1995)

who proposed that the construct of personality has an indirect role in the prediction of

OCB. Importantly the strongest pathway was from interpersonal skills to OCB; which

also had the strongest bivariate relationship. This relationship was demonstrated to be

the most critical in the prediction of OCB for the current study.

Study Considerations, Limitations, and Possible Future Research Directions

The current study aimed to advance knowledge in the prediction of OCB. To

achieve this goal the study required careful consideration of the methodology to

measure the variables that were included in the model. Some of these considerations


have provided an opportunity for additional research. Issues surrounding the ability to

generalise the results due to the main limitations of the study were also considered.

Importantly, the study’s methodology required the development of a new scale –

an interpersonal skills scale. As this was the first research that utilised this scale,

verification would be suggested in additional research to determine its usefulness as a

valid and reliable measure. The five initial items of this scale were based on work of

Ferris et al. (2001). Additional items were added to create a scale of 15 items. As

discussed in Chapter 5, one item was removed during initial analysis resulting in 14

items. Importantly, the interpersonal skills scale of the final 14 items achieved good

reliability for the observer version (0.93) with a slightly lower reliability of 0.80 for the

self-report version. Given that this study provided the first opportunity to utilise the

interpersonal skills scale, further research is required to determine if this scale is

consistent over time.

While test-retest reliability of the interpersonal skills scale is one aspect of the

study that could provide an avenue for additional research, attention could also be

applied to the issue of the interpersonal skills measure not producing the hypothesised

relationship with agreeableness. As discussed, a similar result was found by Ferris et el.

(2001), as measured by their seven-item interpersonal skills scale. The non-significant

bivariate relationship between agreeableness and the observer scores for both

interpersonal skills and OCB in the current study raises the question of why not?

Importantly, Hough and Ones (2001) note that agreeableness is one of the Big Five

dimensions that can possess an unstable factor structure thereby producing inconsistent

results. The characteristics of extreme agreeableness scores may provide some insight

into why this may be the case, as agreeableness may not have a linear relationship with

adaptive behaviour. As discussed in Chapter 3, an extreme high score on agreeableness


can result in maladaptive behaviour (McCrae & Costa, 1987). This previous discussion

suggested that highly agreeable individuals tend to avoid conflict at any costs, which in

some cases does not produce the desired outcomes in a work environment. For example

an employee with extremely high agreeableness may not raise important issues within a

workgroup in an effort to avoid creating conflict. From an external perspective, a lack of

conflict may suggest that the group is working harmoniously. However, the group may

be lacking critical aspects of trust and openness to be able to produce critical analysis. A

lack of criticality can produce maladaptive behaviour within the group creating less

efficiency and effectiveness.

A different type of maladaptive behaviour is known as counterproductive

behaviour, when the individual has an extreme low level of agreeableness. This type of

behaviour can be disruptive in the workplace and has a moderately negative relationship

with OCB (Dalal, 2005). Dalal suggested that agreeableness may have a role in the

prediction of both OCB and counterproductive behaviour. Based on Dalal’s (2005)

suggestion, levels of agreeableness may have a score range that is beneficial to

interpersonal skills or OCB, while extremely positive and negative scores can be

disruptive. The current study did not exclude extreme scores on agreeableness in the

prediction of OCB. It may be the extreme agreeableness scores are related to

counterproductive behaviour in the workplace. However, Dalal acknowledges that this

position is untested and requires further research. Additional research may also involve

the construct of counterproductive behaviour to further understand the antecedents of

OCB and determine what antecedents are common and different for these two important

and related constructs.

In addition to considering the usefulness of agreeableness as a predictor of OCB,

further research could identify other variables that could be included in the model


pathways. The hypothesised model was developed to predict OCB through multiple

variables with a balance between complexity and simplicity. It is important to note that

general mental ability, extraversion, and conscientiousness contributed only to 13

percent of the variance in interpersonal skills. As the explained variance is low,

additional variables could be included in this pathway to increase the explained variance

in interpersonal skills. One such variable is emotional intelligence. Emotional

intelligence is a relatively new construct in the field of Industrial-Organisational

Psychology that has created both interest and controversy (MacCann, 2010). The

construct has been conceptualised as a standard form of intelligence and also as a lower

order trait that encompasses extraversion and neuroticism, which are considered the

emotionally-related facets of personality (Petrides, Pita, & Kokkinaki, 2007). Salovey

and Mayer (1990) define emotional intelligence as an ability to monitor and

discriminate both self and other’s feelings and emotions, then utilise this information to

guide self thinking and actions. In other words, emotional intelligence is the ability to

problem solve and make competent decisions based on both thoughts (logic) and

feelings (intuition) (Mayer & Salovey, 1997). Notably, despite being relatively new, the

construct of emotional intelligence has generated increasing research interest. Of note to

OCB research, Brackett, Rivers, Shiffman, Lener, and Salovey (2006) observed a

positive relationship between emotional intelligence and social functioning, including

social competence and interpersonal sensitivity. Brackett, Rivers, and Salovey (2011)

suggest that emotional intelligence can benefit social functioning through the abilities of

detecting other people’s emotional states, adopting their perspectives, enhancing

communication, and regulating self-behaviour. As such, emotional intelligence could

provide an interesting research direction in the prediction of interpersonal skills.


Other variables that could be included in an expanded OCB prediction model

are neuroticism and negative affect, as discussed in Chapter 3. These variables possess

theoretically based reasons for inclusion in further research. Similarly to the

extraversion–positive affect pathway, the neuroticism–negative affect would be placed

theoretically in the job involvement section of the model. Notably the neuroticism–

negative affect relationship has demonstrated some support for inclusion in previous

studies (e.g. Fogarty, Machin, Albion, Sutherland, Lalor, & Rivitt, 1999), which found

the neuroticism–negative affect had stronger correlations than the extraversion–positive

affect relationship. Additionally negative affect has been associated positively with

counterproductive behaviour and negatively with OCB (Kaplan et al., 2009).

Like all studies, the present research is not without limitations. The main

limitation of the current study was the size of sample. Due to the operational demands

on the Royal Australian Navy at the time of data collection, the proposed sample size

had to be reduced. As previously discussed in Chapter 5, 400 Petty Officers were

targeted initially for inclusion in the study; however, data were only able to be collected

from 202 participants. Additionally, three sets of observer measures were required for

interpersonal skills and OCB. Of the 200 usable data sets, 147 possessed complete data

from all self-report measures and three observer reports. The final sample size of 147 is

a relatively small sample for SEM analysis, as SEM is considered a large sample

statistical technique (Byrne, 2010). The methodology of the current study attempted to

address this issue of small sample size by the utilisation of a goodness-of-fit index that

took small sample size into account, namely the AGFI. Moreover, the current study did

possess an acceptable sample size for the results to be assessed. Nevertheless, small

sample sizes can create less stable solutions and can limit the ability to generalise the


results (Cohen & Manion, 1997). It is recommended that a large sample replication

study is conducted to confirm the results of this study.

A further generalisation issue relates to the sample utilised in the current study.

A military population can be a very homogeneous population. If the sample is

homogeneous, the ability to generalise the results that have been observed to other

populations can be reduced (Cohen & Manion, 1997). Notably this situation is reflective

for the current sample, which collected data from Petty Officers within the Royal

Australian Navy. Schindlmayr and Ong (2001) state that the Australian Defence Force

is under-represented in ethnic uniformed members, as compared to the levels of

ethnicity in Australian society. Moreover, there are a greater percentage of uniformed

males working in the Australian Defence Force than uniformed female members. The

current research studied the rank of Petty Officer, which is a supervisor level. While

ethnicity demographics were not collected as part of this study, it is apparent that the

majority of responses from Petty Officers were collected from males (126 males: 21

females). Furthermore, while the ages of Petty Officers ranged in age from 26 to 57

years, the mean age was 36.26 years with a standard deviation of 5.90 years. However,

it is important to note that while the rank was kept constant for the observed worker, the

area of workplace specialisation varied across 27 different job categories from a Navy

working environment. Considering the possible issues of homogeneity of the targeted

group in terms of ethnicity, age, gender, and work level, it is suggested that the

generalisability of the results be limited to other military populations across different

specialisations at the supervisor level, particularly within the Australian Defence Force.

To increase the generalisability of the results, further research could be conducted to

determine if the same results are replicated. The additional research could utilise

different populations, including public and private industry, as previously mentioned,


and also across levels of supervisor/management, gender, and workforces external to the

military.

The final important methodological issue that needs to be considered in this

study was the use of subjective scales to measure the variables within the tested model.

As noted in Chapter 3, subjective scales are based on judgments, feelings, and human

experience (Muckler & Seven, 1992) and have the potential to produce bias affects such

as social desirable responses and common method bias. However, it is acknowledged

that subjective scales are superior to objective scales when measuring the variables to

predict OCB. Specifically, Muckler and Seven argue that subjective scales include what

is happening, past experience, present knowledge, intentions, and motivational levels to

provide a richer and more unique measurement of human behaviour than objective

measures. However, when considering the potential issues associated with subjective

scales, I selected where possible established measures that had previously demonstrated

good validity and reliability. These scales included the NEO PI-R (Costa & McCrae,

1992a), Raven’s Progressive Matrices (Raven et al., 1998), PANAS (Watson et al.,

1988), JIS (Lodahl & Kejnar, 1965), MSQ (Weiss et al., 1967), and OCB scale

(Podsakoff et al., 1990). Moreover, the methodology included both self-report scales for

personality, affective disposition, job satisfaction, job involvement, general mental

ability; while interpersonal skills and OCB were measured using a 360 degree strategy

to reduce the effect of common method bias, as recommended by Organ et al. (2006).

As such, the potential problems associated with subjective scales were minimised.

Practical Significance of the Research

OCB is a specific type of job performance and the term was first used by

Bateman and Organ in 1983. Since its inception there have been a number of definitions


with the most recent being “individual behavior that is discretionary, not directly or

explicitly recognized by the formal reward system, and in aggregate promotes the

efficient and effective functioning of the organization” (Organ et al., 2006, p.3). The

discretionary nature of OCB suggests an unstated assumption that OCB can be

produced by any individual who chooses to do so (Dudley & Cortina, 2008). This point

raises two important questions when considering the results of the current study. Firstly,

if OCB is discretionary, what contribution, if any, can individual differences based on

traits have on the production of OCB? Secondly, if OCB has a strong relationship with

interpersonal skills, what are the competency requirements to produce OCB? Model

testing of the OCB prediction attributes help to address the first question. Specifically,

the current study has provided further evidence to suggest that general mental ability,

personality (conscientiousness and extraversion), and affective disposition (positive

affect) are involved in the prediction of OCB, albeit as small indirect contributions. As

such, individuals who possess high levels of these traits may produce more OCB than

individuals with lower scores.

Results that support the role of individual differences in the prediction of OCB

can assist in the development of strategies for personnel recruitment and selection.

While it seems appropriate to select personnel based on a match of the traits of potential

candidates to current personnel, it is also noted that there are advantages and

disadvantages to this human resource strategy. For example, hiring candidates with

genetically based mood tendencies can provide both beneficial and non-beneficial

outcomes to the organisation. Affect has been shown to impact memory via attention

and recall processes, being stronger and more general for positive affect states within

normal populations (Moore & Isen, 1990). Additionally, Ilies and Judge (2002) found

that variability in mood can influence both the level and variance of job satisfaction,


with the effects of mood fluctuation being more consistent for positive affect than for

negative affect (Moore & Isen, 1990). Importantly, Judge (1993) noted that an

employee with high positive affect but low vocational interest is likely to have a higher

turnover rate as this type of person is motivated to gain a successful personal outcome

and will be more proactive to achieve this aim. Moreover, while affective disposition

can be considered a trait, mood variations can vary considerably over time.

Klehe (2004) acknowledges that while selection processes need to be valid, they

also should take into account the internal and external pressures that occur within the

organisation. Hiring the same type of person can reduce the diversity within the

workplace and also aspects of innovation. Notably, Muchinsky (1999) noted that teams

comprised solely of extraverted individuals can reduce the effectiveness of the team as a

product of the internal dynamics. Given that workplace behaviour is complex and often

situation driven, flexibility and diversity can provide business with a competitive edge.

The second question of producing OCB in the workplace considers the issue of

competency. Werner (2000) noted that this issue has been mostly considered from a

recruitment and selection perspective. If OCB is required, then the company could

select individuals who are pre-disposed and/or already possess these skills. Notably the

current study supported the role of interpersonal skills as a mediator for both general

mental ability and extraversion to predict OCB. Specifically, the results demonstrated a

very high bivariate correlation between interpersonal skills and OCB (r=0.81). This

would support the assumption that some individuals are pre-disposed to produce higher

levels of OCB than individuals who have lower levels of these traits. However, while

the dispositions of general mental ability, conscientiousness, and extraversion possessed

direct pathways to interpersonal skills, their pathway coefficients were small to

moderate (0.19, 0.25, and 0.32, respectively). This suggests that these stable traits play a


small casual role to produce interpersonal skills. Additionally, only 13 percent of the

variance was explained in interpersonal skills based on the general mental ability,

conscientiousness, and extraversion pathways. The results suggested that while some

aspects of interpersonal skills could be stable and non-changeable due to dispositional

influences, other aspects may be able to be modified. These results have important

implications for the training of interpersonal skills and subsequent increase in the

production of OCB in the workplace. For an individual to be trained in the area of

interpersonal skills, they must have the ability to change their behaviour through

learning. The construct of interpersonal skills must contain aspects that are changeable

through training and not purely caused by stable traits. Notably stable and non-

malleable traits actually impede training. With the possibility that interpersonal skills

possessed aspects of changeability, training becomes a possibility. It may be possible to

develop the individual’s interpersonal skills to increase their competency in OCB.

The possibility of developing an individual’s interpersonal skills provides useful

information for the area of training and education (Werner, 2000). If a person is

demonstrating low levels of OCB in the workplace, it may be possible as part of their

development plan to receive interpersonal skills training to improve their OCB

performance. One only has to conduct a Google search of interpersonal skills training

on the internet to see that there is a plethora of service providers conducting

interpersonal skills training in areas that include but are not limited to: team building,

communication, conflict resolution, negotiation, and networking. Given the results of

the current study, it may be fair to assume that individual differences (general mental

ability, extraversion, and conscientiousness) could assist in the production of OCB

through interpersonal skills. This issue would provide an interesting area of further

research. Additionally as mentioned, positive affect, job satisfaction, and job


involvement are important in the role of OCB production. These aspects would also be

useful to consider in future study.

This study has contributed to the continued effort to understand the prediction of

OCB. Importantly, the current study has demonstrated that the prediction of OCB

involves more than just job satisfaction and personality. The results have indicated that

interpersonal skills are important for producing OCB in the workplace. Additionally,

motivational aspects contribute to the resultant OCB, with less influence than

interpersonal skills. Specifically, the model indicates that to select pre-disposed highly

motivated personnel, an applicant should score highly on extraversion,

conscientiousness, and positive affect.

OCB has been identified as an important behaviour within the workplace as it

can directly and indirectly affect the bottom line (Organ et al., 2006). Consequently, it is

not surprising that the OCB construct has been widely investigated. Organisations must

stay competitive in order to retain and develop their position in the market place. One

way this is achieved is through ensuring that their personnel can competently carry out

their daily tasking. This goal can be enhanced through appropriate selection and

development strategies. Predictive personnel research remains critical for Industrial-

Organisational psychologists to develop accurate and cost effective personnel strategies.

Given the findings of the current study in conjunction with previous research,

organisations can further develop their strategies to select and develop personnel based

on OCB related dispositions and learning capabilities with a balance to task

performance requirements.


REFERENCES

Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in

Experimental Psychology (Vol. 2, pp. 267-299). New York: Academic Press.

Aiken, L.R. (1993). Personality: Theories, Research, and Applications. Englewood

Cliffs, NJ: Prentice Hall.

Allen, T. D., Barnard, S., Rush, M. C., & Russell, J. E. A. (2000). Ratings of

organizational citizenship behavior: Does the source make a difference? Human

Resource Management Review, 10(1), 97-114.

Allport. G.W. (1961). Pattern and Growth in Personality. New York: Holt, Rinehart &

Winston.

Allport, G.W. & Odbert, H.S. (1936). Trait names: A psycho-lexical study.

Psychological Monograms, 47, (211).

Arbuckle, J.L. (2008). AMOS 17.0.0. (Build 1404). Crawfordville, FL: AMOS

Development Corporation.

Arbuckle, J.L. & Wothke, W. (1999). AMOS 4.0 User’s Guide. Chicago, IL: SPSS Inc,

Small Waters Corporation.

Argyle, M. (1969). Social Interaction. Chicago: Aldine.

Argyle, M. (1994). The psychology of interpersonal behaviour, (5th Ed.). London:

Penguin.

Arnold, J., Cooper, C.L., & Robertson, I.T. (1998). Work Psychology: Understanding

Human Behaviour in the Workplace, (3rd

Ed.). London: Financial Times

Management.

Atkinson, R.L., Atkinson, R.C., Smith, E.E., & Bem, D.J. (1993). Introduction to

Psychology. (11th

ed.). Fort Worth: Harcourt Brace College.

Baird, K., Harrison, G., & Reeve, R. (2007). The culture of Australian organizations

and its relation with strategy, International Journal of Business Studies, 15(1), 15-

41.

Baron, R.M. & Kenny, D.A. (1986). The moderator-mediator variable distinction in

social psychological research: Conceptual, strategic, and statistical considerations.

Journal of Personality and Social Psychology, 51(6), 1173-1182.

Barnard, C. I. (1938). The Functions of the Executive. Cambridge, MA: Harvard

University Press.

Barrick, M.R. & Mount, M.K. (1991). The big five personality dimensions and job

performance: A meta-analysis, Personnel Psychology, 44(1), 1-26.


Barrick, M.R. & Mount, M.K. (2005). Yes, personality matters: Moving on to more

important matters. Human Performance, 18(4), 359-372.

Barrick, M.R., Mount, M.K., & Strauss, J.P. (1993). Conscientiousness and

performance of sales representative: Test of the mediating effects of goal setting.

Journal of Applied Psychology, 78(5), 715-722.

Barrick, M.R., Stewart, G.L., & Piotrowski, M. (2002). Personality and job

performance: Test of the mediating effects of motivation among sales

representatives. Journal of Applied Psychology, 87(1), 43-51.

Bateman, T. S. & Organ, D. W. (1983). Job satisfaction and the good soldier: The

relationship between affect and employee “citizenship”. Academy of Management

Journal, 26, 587-595.

Bentler, P.M. (2005). EQS Structural Equations Program Manual. Encino, CA:

Multivariate Software.

Blau, P. M. (1964). Exchange and Power in Social Life. New York: Wiley.

Blunch, N.J. (2008). Introduction to Structural Equation Modelling Using SPSS and

AMOS. Thousand Oaks, CA: SAGE.

Borman, W.C. (1997). 360˚ ratings: An analysis of assumptions and a research agenda

for evaluating their validity. Human Resource Management Review, 7(3), 299-

315.

Borman, W.C. (2004). The concept of organizational citizenship. Current Directions in

Psychological Science, 13(6), 238-241.

Borman, W. C. & Motowidlo, S. J. (1993). Expanding the criterion domain to include

elements of contextual performance. In N. Schmitt & W.C. Borman (Eds.),

Personnel Selection in Organizations. San Francisco: Jossey-Bass.

Borman, W. C. & Motowidlo, S. J. (1997). Task performance and contextual

performance: The meaning for personnel selection research. Human Performance,

10(2), 99-109.

Borman, W.C., White, L.A., & Dorsey, D.W. (1995). Effects of ratee task performance

and interpersonal factors of supervisor and peer performance ratings. Journal of

Applied Psychology, 80(1), 168-177.

Bowling, N.A. (2007). Is the job satisfaction-job performance relationship spurious? A

meta-analytic examination. Journal of Vocational Behavior, 71(2), 167-185.

Boyd, B. & Lees, R. (2012). Australia. In W. Kirsten & R.C. Karch (Eds.), Global

Perspectives in Workplace Health Promotion, (pp. 1-20). Sudbury, MA: Jones &

Bartlett Learning, LLC.

Bracken, D.W., Timmreck, C.W., & Church, A.H. (2001). Introduction: A multisource

feedback process model. In D.W. Bracken, C.W. Timmreck, & A.H. Church

(Eds.), The Handbook of Multisource Feedback (pp.3-14). San Francisco, CA:

Jossey-Bass, Inc.


Brackett, M.A., Rivers, S.E., & Salovey, P. (2011). Emotional intelligence: Implications

for personal, social, academic, and workplace success. Social and Personality

Psychology Compass, 5(1), 88-103.

Brackett, M.A., Rivers, S.E., Shiffman, S., Lerner, N., & Salovey, P. (2006). Relating

emotional abilities to social functioning: A comparison of self-report and

performance measures of emotional intelligence. Journal of Personality and

Social Psychology, 91(4), 780-795.

Brief, A.P. & Motowidlo, S.J. (1986). Prosocial organizational behaviors. Academy of

Management Review, 11(4), 710-725.

Brown, S.P. (1996). A meta-analysis and review of organizational research on job

involvement. Psychological Bulletin, 120(2), 235-255.

Brown, S.P. & Leigh, T.W. (1996). A new look at psychological climate and its

relationship to job involvement, effort, and performance. Journal of Applied

Psychology, 81(4), 358-368.

Browne, M.W. & Cudeck, R. (1993). Alternative ways of assessing model fit. In K.A.

Bollen & J.S. Long (Eds.), Testing Structural Equation Models (136-162).

Newbury Park, CA: Sage.

Byrne, B.M. (2010). Structural Equation Modeling with AMOS: Basic Concepts,

Applications, and Programming, (2nd

ed.). New York, NY: Routledge.

Campbell, J.P. (1990). Modeling the performance prediction problem in industrial

organizational psychology, (Chapter 12, 689-732). In M.D. Dunnette & L.M.

Hough (eds.), Handbook of Industrial and Organizational Psychology. Palo Alto,

CA: Consulting Psychologists Press.

Campbell, J. P., McCloy, R. A., Oppler, S. H., & Sager, C. E. (1993). A theory of

performance. In N. Schmitt & W.C. Borman (Eds.), Personnel Selection in

Organizations. San Francisco, CA: Jossey-Bass.

Cattell, R.B. (1957). Personality and Motivation Structure and Measurement. New

York, NY: World Book Company.

Cohen, L. & Manion, L. (1997). Research Methods in Education, (4th

ed.). London:

Routledge.

Conway, J.M. & Huffcutt, A.I. (1997). Psychometric properties of multisource

performance ratings: A meta-analysis of subordinate, supervisor, peer, and self-

ratings. Human Performance, 10(4), 331-360.

Cook, J.D., Hepworth, S.J., Wall, T.D., & Warr, P.B. (1981). The Experience of Work:

A Compendium and Review of 249 Measures and Their Use. London: Academic

Press.

Cook, H.E. & Manson, G.E. (1926). Abilities necessary in effective retail selling and a

method of evaluating them. Journal of Personnel Research, 5(000001), 74-82.


Costa, P.T., Jr., & McCrae, R.R. (1980). Influence of extraversion and neuroticism on

subjective well-being: Happy and unhappy people. Journal of Personality and

Social Psychology, 38(4), 668-678.

Costa, P.T., Jr., & McCrae, R.R. (1992a). NEO-PI-R Professional Manual. Florida:

Psychological Assessment Resources, Inc.

Costa, P.T. Jr & McCrae, R.R. (1992b). NEO Five-Factor Inventory (NEO-FFI)

Professional Manual. Odessa, FL: Psychological Assessment Resources, Inc.

Costa, P.T. & McCrae, R.R. (1995). Solid ground in the wetlands of personality: A

reply to Block. Psychological Bulletin, 111(2), 216-220.

Craig, S.B. & Hannum, K. (2006). Research update – 360-Degree performance

assessment. Consulting Psychology Journal: Practice and Research, 58(2), 117-

122.

Dalal, R.S. (2005). A meta-analysis of the relationship between organizational

citizenship behavior and counterproductive work behavior. Journal of Applied

Psychology, 90(6), 1241-1255.

Dawis, R.V. (2001). 1999 Leona Tyler award: Rene V. Dawis: Toward a psychology of

values. The Counseling Psychologist, 29(3), 458.

Department of Defence. (2001). Defence Instruction (Navy) PERS 2-2: Instruction for

Category Sponsors. Canberra, ACT: Department of Defence.

Diefendorff, J., Brown, D.J, Kamin, A. & Lord, B. (2002), ‘Examining the roles of job

involvement and work centrality in predicting organizational citizenship behaviors

and job performance. Journal of Organizational Behavior, 23(1), 93-108.

Digman, J.M. (1989). Five robust trait dimensions: Development, stability, and utility.

Journal of Personality, 57(2), 195-214.

Digman, J.M. (1996). The curious history of the five-factor model. In J.S. Wiggins (Ed.)

The Five-Factor Model of Personality: Theoretical Perspectives (pp. 1–20). New

York, NY: Guilford Press.

Doverspike, D., Cober, A. B., & Arthur, W. Jr. (2004). Multi-aptitude test batteries. In

J.C. Thomas (Ed.), Comprehensive Handbook of Psychological Assessment: Vol.

4, Industrial and Organizational Assessment, 35-55. New Jersey: John Wiley &

Sons, Inc.

Dudley, N.M. & Cortina, J.M. (2008). Knowledge and skills that facilitate the personal

support dimension of citizenship. Journal of Applied Psychology, 93(6), 1249-

1270.

Epstein, S.P. (1983). The stability of confusion: A reply to Mischel and Peake.

Psychological Review, 90(2), 179-184.

Eysenck, H.J. (1967). The Biological Basis of Personality. Springfield, IC: Charles C

Thomas.


Eysenck, H.J. (1982). Personality, Genetics and Behavior: Selected Papers. New York:

Praeger.

Eysenck, H.J. (1998). Dimensions of Personality. New Brunswick, New Jersey:

Transaction Publishers. (Original: 1947).

Eysenck, H.J. & Eysenck, M.W. (1985). Personality and Individual Differences: A

Natural Science Approach. New York, NY: Plenum Publishing Corporation

Fay, D. & Sonnentag, S. (2010). A look back to move ahead: New directions for

research on proactive performance and other discretionary work behaviours.


Ferris, G.R., Witt, L.A., & Hochwarter, W.A. (2001). Interaction of social skill and

general mental ability on job performance and salary. Journal of Applied

Psychology, 86(6), 1075-1082.

Feshbach, S. & Weiner, B. (1991). Personality. (3rd

Ed.). Lexington, MA: D.C. Health

and Company.

Field, A. (2009). Discovering Statistics Using SPSS (3rd

Ed.). Thousand Oaks, CA:

SAGE.

Fisher, C.D. (2000). Mood and emotion while working: Missing pieces of job

satisfaction? Journal of Organizational Behavior, 21(2), 185-202.

Fleenor, J.W., Smither, J.W., Atwater, L.E., Braddy, P.W., & Sturm, R.E. (2010). Self-

other rating agreement in leadership: A review. The Leadership Quarterly, 21(6),

1005-1034.

Fletcher, C. & Baldy, C. (2000). A study of individual differences and self-awareness in

the context of multi-source feedback. Journal of Occupational and

Organizational Psychology, 73(3), 303-319.

Fletcher, C., Baldy, C., Cunningham-Snell, N. (1998). The psychometric properties of

360 degree feedback: An empirical study and a cautionary tale. International

Journal of Selection and Assessment, 6(1), 19-34.

Fogarty, G.J., Machin, M.A., Albion, M.J., Sutherland, L.F., Lalor, G.I., & Rivitt, S.

(1999). Predicting occupational strain and job satisfaction: The role of stress,

coping, personality, and affectivity variables. Journal of Vocational Behavior,

54(3), 429-452.

Gardner, H. (1993). Multiple Intelligence: The Theory in Practice. New York, NY:

Basic Books.

Gatewood, R.D. & Field, H.S. (1998). Human Resource Selection, (4th

ed.). New York,

NY: The Dryden Press.

Gellatly, I.R. (1996). Conscientiousness and task performance: Test of a cognitive

process model. Journal of Applied Psychology, 81(5), 474-482.


George, J.M. (1990). Personality, affect, and behavior in groups. Journal of Applied

Psychology, 75(2), 107-116.

George, J.M. (1991). State or trait: Effects of positive mood on prosocial behaviors at

work. Journal of Applied Psychology, 76(2), 299-307.

George, J.M. (1992). The role of personality in organizational life: Issues and evidence.

Journal of Management, 18(2), 185-213.

George, J.M. & Bettenhausen, K. (1990). Understanding prosocial behavior, sales

performance, and turnover: A group level analysis in a service context. Journal of


George, J.M. & Brief, A.P. (1992). Feeling good – doing good: A conceptual analysis of

the mood at work - organizational spontaneity relationship. Psychological

Bulletin, 112(2), 310-329.

George, J.M. & Jones, G.R. (1996). The experience of work and turnover intentions:

interactive effects of value attainment, job satisfaction, and positive mood.


George, J.M. & Jones, G.R. (1997). Organizational spontaneity in context. Human

performance, 10(2), 153-170.

Goldberg, L.R. (1990). An alternative ‘description of personality’: The big-five factor

structure. Journal of Personality and Social Psychology, 59(6), 1216-1229.

Graham, J.W. (1991). An essay on organizational citizenship behavior. Employee

Responsibilities and Rights Journal, 4(4), 249-270.

Grubb III, W. L., Whetzel, D. L., & McDaniel, M. A. (2004). General mental ability

tests in industry. In J.C. Thomas (Ed.), Comprehensive Handbook of

Psychological Assessment: Vol. 4, Industrial and Organizational Assessment, 7–

21. New Jersey: John Wiley & Sons, Inc.

Halfhill, T.R., Neilsen, T.M., & Sunstrom, E. (2008). The ASA framework: A field

study of group personality composition and group performance in military action

teams. Small Group Research, 39(5), 616-635.

Hannam, R. & Jimmieson, N. (2002). The relationship between extra-role behaviours

and job burnout for primary school teachers: A preliminary model and

development of an organisational citizenship behaviour scale. Unpublished paper.

Australia: University of Queensland. Available Online:

http://www.Aare.edu.au/02pap/han02173.htm.

Hargie, O., Saunders, C., & Dickson, D. (1994). Social Skills in Interpersonal

Communication, (3rd

Ed.). New York, NY: Routledge.

Harris, M.M. & Schaubroeck, J. (1988). A meta-analysis of self-boss, self-peer, and

peer-supervisor ratings. Personnel Psychology, 41(1), 43-62.

Hayes, J. (2002). Interpersonal Skills at Work (2nd

ed.). New York, NY: Routledge.


Hedge, J.W., Borman, W.C., & Birkeland, S.A. (2001). History and development of

multisource feedback as a methodology. In D.W. Bracken, C.W. Timmreck, &

A.H. Church (Eds.). The Handbook of Multisource Feedback, (pp. 15-32). San

Francisco, CA: Jossey-Bass Inc.

Hochwarter, W.A., Perrewé, P.L., Ferris, G.R., & Brymer, R.A. (1999). Job satisfaction

and performance: The moderating effects of value attainment and affective

disposition. Journal of Vocational Behavior, 54(2), 296-313.

Hoffman, B.J., Blair, C.A., Meriac, J.P., & Woehr, D.J. (2007). Expanding the

criterion? A quantitative review of the OCB literature. Journal of Applied

Psychology, 92(2), 555-566.

Hogan, R. (1983). A socioanalytic theory of personality. In M. M. Page (Ed.), 1982

Nebraska Symposium on Motivation (pp. 55–89). Lincoln: University of Nebraska

Press.

Hogan, R. (1991). Personality and personality measurement. In Dunnette, M.D. & L.M.

Hough (Eds), Handbook of Industrial and Organizational Psychology: Vol 2, 873-

919. Palo Alto, CA: Consulting Psychologists Press.

Hogan, J. & Holland, B. (2003). Using theory to evaluate personality and job-

performance relations: A socioanalytic perspective. Journal of Applied

Psychology, 88(1), 100-112.

Hough, L.M. (1992). The “big five” personality variables – Construct confusion:

Description versus prediction. Human Performance, 5(1&2), 139-155.

Hough, L.M.. & Ones, D.S. (2001). The structure, measurement, validity, and use of

personality variables in industrial, work, and organizational psychology. In N.

Anderson, D.S. Ones, H.K. Sinangil, and C. Viswesvaran (Eds.), Handbook of

Industrial, Work, and Organizational Psychology, Vol. 1: Personnel psychology.

New York: Wiley.

Hu, L-T. & Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure

analysis: Conventional criteria versus new alternatives. Structural Equation

Modeling, 6(1), 1-55.

Hui, C., Law, K.S., & Chen, Z.X. (1999). A structural equation model of the effects of

negative affectivity, leader-member exchange, and perceived job mobility on in-

role and extra-role performance: A Chinese case. Organizational Behavior and

Human Decision Processes, 77(1), 3-21.

Iaffaldano, M.T. & Muchinsky, P.M. (1985). Job satisfaction and job performance: A

meta-analysis. Psychological Bulletin, 97(2), 251-273.

Ilies, R. & Judge, T.A. (2002). Understanding the dynamic relationships among

personality, mood, and job satisfaction: A field experience sampling study.

Organizational Behavior and Human Decision Processes, 89(2), 1119-1139.

Ivancevich, J.M., Olekalns, M.T., & Matteson, M. (1997). Organisational Behaviour

and Management. Sydney: Irwin.


Jensen-Campbell, L.A. & Graziano, W.G. (2001). Agreeableness as a moderator of

interpersonal conflict. Journal of Personality, 69(2), 323-362.

John, O.P. (1990). The “big five” factor taxonomy: Dimensions of personality in the

natural language and in questionnaires. In L.A. Pervin (Ed.), Handbook of

Personality: Theory and Research. New York/London: Guilford Press.

Johnson, K. & Jacka, E. (2005). 2004 Australian Defence Force Exit Survey Report

Reasons for Leaving. (Directorate of Strategic Personnel Planning and Research

DSPPR Research Report 26/2005). Canberra, ACT: Commonwealth of Australia.

Jöreskog, K.G. (1993). Testing structural equation models. In K.A. Bollen & J.S. Long

(Eds.), Testing Structural Equation Models (pp.294-316). Newbury Park, CA:

Sage.

Judge, T.A. (1993). Does affective disposition moderate the relationship between job

satisfaction and voluntary turnover? Journal of Applied Psychology, 78(3), 395-

401.

Judge, T.A., Thoresen, C.J., Bono, J.E., & Patton, G.K. (2001). The job satisfaction-job

performance relationship: A qualitative and quantitative review. Psychological

Bulletin, 127(3), 376-407.

Kaplan, R.M. & Saccuzzo, D.P. (1997). Psychological Testing: Principles,

Applications, and Issues (4th

ed.). International Edition: Brooks/Cole.

Kaplan, S., Bradley, J.C., Luchman, J.N., & Haynes, D. (2009). On the role of positive

and negative affectivity in job performance: A meta-analytic investigation.


Katz, D. (1964). The motivational basis of organizational behavior. Behavioral

Sciences, 9(2), 131-46.

Katz, D. & Kahn, R. L. (1966). The Social Psychology of Organizations. New York:

Wiley.

Katz, D. & Kahn, R. L. (1978). The Social Psychology of Organizations (2nd

Ed.). New

York: Wiley.

Katzell, R.A. & Thompson, D.E. (1990). Work motivation: Theory and practice.

American Psychologist, 45(2), 144-153.

Klehe, U.C. (2004). Choosing how to choose: Institutional pressures affecting the

adoption of personnel selection procedures. International Journal of Selection and

Assessment, 12(4), 327-342.

Kline, P. (2003). The New Psychometrics: Science, Psychology and Measurement. New

York, NY: Routledge.

Kline, R.B. (2005). Principles and Practice of Structural Equation Modeling (2nd ed.).

New York: Guilford.


Kline, R.B. (2011). Principles and Practice of Structural Equation Modeling (3rd ed.).

New York: Guilford.

Kline, T.J.B. & Sulsky, L.M. (2009). Measurement and assessment issues in

performance appraisal. Canadian Psychology, 50(3), 161-171.

LePine, J.A., Erez, A., & Johnson, D.E. (2002). The nature and dimensionality of

organizational citizenship behavior: A critical review and meta-analysis. Journal

of Applied Psychology, 87(1), 52-65.

Locke, E.A. (1976). The nature and causes of job satisfaction. In M.D. Dunnette (Ed.),

Handbook of Industrial and Organizational Psychology (Section VI, Chapter 30,

1297-1349). Chicago, IL: Rand McNally.

Lodahl, T.M. & Kejnar, M. (1965). The definition and measurement of job

involvement. Journal of Applied Psychology, 49(1), 24-33.

Lucas, R.E., Le, K., & Dyrenforth, P.S. (2008). Explaining the extraversion/positive

affect relation: Sociability cannot account for extraverts’ greater happiness.


Luthans, K.W. & Farner, S. (2002). Expatriate development – The use of 360-degree

feedback. Journal of Management Development, 21(10), 780-793.

MacCann, (2010). Further examination of emotional intelligence as a standard

intelligence: A latent variable analysis of fluid intelligence, crystallized

intelligence, and emotional intelligence. Personality and Individual Differences,

49(5), 490-496.

Maddi, S.R. (2001). Personality Theories: A Comparative Analysis. (6th

Ed.). Prospect

Heights, IL: Waveland Press.

Marinova, S.V., Moon, H., & Van Dyne, L. (2010). Are all good soldier behaviors the

same? Supporting multidimensionality of organizational citizenship behaviors

based on rewards and roles. Human Relations, 63(10), 1463-1485.

Matthews, G., Deary, I.J., & Whiteman, M.C. (2009) Personality Traits (3rd Ed).

Cambridge: Cambridge University Press.

Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D.

J. Sluyter (Eds.), Emotional Development and Emotional Intelligence:

Educational Implications (pp. 3–34). New York, NY: Basic Books, Inc.

McAdams, D.P. (1992). The five-factor model in personality: A critical appraisal.


McCloy, R.A., Campbell, J.P., & Cudeck, R. (1994). A confirmatory test of a model of

performance determinants. Journal of Applied Psychology, 79(4), 493-505.

McCrae, R.R. & Costa, P.T. (1985) Updating Norman's "adequate taxonomy":

Intelligence and personality dimensions in natural language and in questionnaires,

Journal of Personality & Social Psychology, 49(3), 710-721.


McCrae, R.R. & Costa, P.T. (1987). Validation of the five factor model of personality

across instruments and observers. Journal of Personality and Social Psychology,

52(1), 81-90.

McCrae, R.R. & John, O.P. (1992). An introduction to the five-factor model and its

applications. Journal of Personality, 60(2), 175-215.

McCrae, R.R. & Costa, P.T. (1991). Adding Liebe und Arbeit: The Full Five-Factor

Model and Well-Being. Personality and Social Psychology Bulletin, 17(2), 269-

273.

McKinnon, A., Montalban, M., Barton, N., & Gloyn, E. (2005). Defence Attitude

Survey 2004 Results Report. (Directorate of Strategic Personnel Planning and

Research DSPPR Research Report 8/2005). Canberra, ACT: Commonwealth of

Australia.

Meichebbaum, D., Butler, L., & Gruson, L. (1981). Toward a conceptual model of

social competence. In J. Wine & D. Smye (Eds.), Social Competence, 36-60. New

York: Guilford Press.

Mitchell, T.R. (1997). Matching motivational strategies with organizational contexts. In

B.M. Staw & L.L. Cummings (Eds.), Research in Organizational Behavior (Vol.

19, 57-149). Greenwich, CT: JAI Press.

Moore, B.S. & Isen, A.M. (1990). Affect and social behavior. In B.S. Moore & A.M.

Isen (Eds.), Affect and Social Behavior (Chapter 1, 1-21). Cambridge: Cambridge

University Press.

Moorman, R.H. & Blakely, G.L. (1995). Individualism-collectivism as an individual

difference predictor of organizational citizenship behavior. Journal of

Organizational Behavior, 16(2), 127-142.

Moorman, R.H., Blakely, G.L., & Niehoff, B.P. (1998). Does perceived organizational

support mediate the relationship between procedural justice and organizational

citizenship behavior? Academy of Management Journal, 41(3), 351-357.

Morgeson, F.P., Mumford, T.V., & Campion, M.A. (2005). Coming full circle: Using

research to address 27 questions about 360-degree feedback programs. Consulting

Psychology Journal: Practice and Research, 57(3), 196-209.

Motowidlo, S.J., Borman, W.C., & Schmit, M.J. (1997). A theory of individual

differences in task and contextual performance. Human Performance, 10(2), 71-

83.

Motowidlo, S.J. & Van Scotter, J.R. (1994). Evidence that task performance should be

distinguished from contextual performance. Journal of Applied Psychology, 79(4),

475-480.

Mount, M. K., & Barrick, M. R. (1998). Five reasons why the “Big Five” article has

been frequently cited. Personnel Psychology, 51(4), 849-857.


Mount, M.K., Barrick, M.R., & Stewart, G.L. (1998a). Five-factor model of personality

and performance in jobs involving interpersonal interactions. Human

Performance, 11(2/3), 145-165.

Mount, M.K., Judge, T.A., Scullen, S.E., Sytsma, M.R., & Hezlett, S.A. (1998b). Trait,

rater and level effects in 360-degree performance ratings. Personnel Psychology,

51(3), 557-576.

Mowday, R.T., Steers, R,M., & Porter, L.W. (1979). The measurement of

organizational commitment. Journal of Vocational Behavior, 14(2), 224-247.

Muchinsky, P.M. (1999). Psychology Applied to Work: An Introduction to Industrial

and Organizational Psychology, (6th

ed.). Belmont, CA: Wadsworth/Thomson

Learning.

Muckler, F.A. & Seven, S.A. (1992). Selecting performance measures: “Objective”

versus “subjective” measurement. Human Factors, 34(4), 441-445.

Nemanick, R.C. Jr., & Munz, D.C. (1997). Extraversion and neuroticism, trait mood,

and state affect: A hierarchical relationship. Journal of Social Behavior and

Personality, 12(4), 1079-1092.

Neuman, G.A., Wagner, S.H., & Christiansen, N.D. (1999). The relationship between

work-team personality composition and the job performance of teams. Group

Organization Management, 24(1), 28-45.

Neuman, G.A. & Wright, J. (1999). Team effectiveness: Beyond skills and cognitive

ability. Journal of Applied Psychology, 84(3), 376-389.

Norman, W.T. (1963). Toward an adequate taxonomy of personality attributes:

Replicated factor structure in peer nomination personality ratings', Journal of

Abnormal & Social Psychology, 66(6), 574-583.

O’Reilley, C. & Chatman, J. (1986). Organizational commitment and psychological

attachment: The effects of compliance, identification, and internalization on

prosocial behavior. Journal of Applied Psychology, 71(3), 492-499.

Organ, D.W. (1977). A reappraisal and re-interpretation of the satisfaction-causes-

performance hypothesis. Academy of Management Review, 2, 46-53.

Organ, D.W. (1988). Organizational citizenship behavior: The good soldier syndrome.

Lexington, MA: Lexington Books.

Organ, D.W. (1990a). The motivational basis of organizational citizenship behavior. In

B.M. Staw & L.L. Cummings (Eds.), Research in Organizational Behavior,

(Volume 12, pp. 43-72). Greenwich, CT: JAI Press.

Organ, D.W. (1990b).The subtle significance of job satisfaction. Clinical Laboratory

Management Review, 4(1), 94-98.

Organ, D.W. (1994). Personality and organizational citizenship behavior. Journal of

Management, 20(2), 465-478.


Organ, D.W. (1997). Organizational citizenship behavior: It’s construct clean-up time.

Human Performance, 10(2), 85-97.

Organ, D.W. & McFall, J.B. (2004). Personality and citizenship behavior in

organizations. In B. Schneider & D.B. Smith (Eds.), Personality and

Organizations, (pp. 291-316). Mahwah, NJ: Lawrence, Erlbaum.

Organ, D.W., Podsakoff, P. M., & MacKenzie, S. B. (2006). Organizational citizenship

behavior: Its nature, antecedents, and consequences. Thousand Oaks, CA: Sage

Publications.

Organ, D.W. & Ryan, K. (1995). A meta-analytic review of attitudinal and dispositional

predictors or organizational citizenship behavior. Personnel Psychology, 48(4),

775-802.

Orr, J.M., Sackett, P.R., & Mercer, M. (1989). The role of prescribed and non-

prescribed behaviors in estimating the dollar value of performance. Journal of


Paullay, I.M, Alliger, G.M., & Stone-Romero, E.F. (1994). Construct validation of two

instruments designed to measure job involvement and work centrality. Journal of


Payne, T. (1998). Editorial – 360 Degree Assessment and Feedback. International

Journal of Selection and Assessment, 6(1), 16-18.

Pervin, L. A. (1990). A brief history of modern personality theory. In L.A. Pervin (Ed.),

Handbook of Personality – Theory and Research, (pp.3-18). New York, NY:

Guilford Press.

Petrides, K.V., Pita, R., & Kokkinaki, (2007). The location of trait emotional

intelligence in personality factor space. British Journal of Psychology, 98(2), 273-

289.

Podsakoff, P.M., MacKenzie, S.B., Paine, J.B., & Bachrach, D.G. (2000).

Organizational citizenship behaviors: A critical review of the theoretical and

empirical literature and suggestions for future research. Journal of Management,

26(3), 513-563.

Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., & Podsakoff, N.P. (2003). Common

method biases in behavioral research: A critical review of the literature and

recommended remedies. Journal of Applied Psychology, 88(5), 879-903.

Podsakoff, P.M., MacKenzie, S.B., Moorman, R.H., & Fetter, R. (1990).

Transformational leader behaviors and their effects on followers’ trust in leader,

satisfaction, and organizational citizenship behaviors. Leadership Quarterly, 1(2),

107-142.

Podsakoff, P. M. & Organ, D. W. (1986). Self-reports in organizational research:

Problems and prospects. Journal of Management, 12(4), 531-544.


Podsakoff, P.M. & Williams, L.J. (1986). The relationship between job performance

and job satisfaction. In E.A. Locke (Ed.), Generalizing for the Laboratory to Field

Settings, Lexington: Lexington Books.

Power, M. & Nottage, C. (2005). 2004 Australian Defence Force Reserves Attitude

Survey Report. (Directorate of Strategic Personnel Planning and Research DSPPR

Research Report 5/2005). Canberra, ACT: Commonwealth of Australia.

Puffer, S.M. (1987). Pro-social behavior, non-compliant behavior and work

performance among commission salespeople. Journal of Applied Psychology,

72(4), 615-621.

Raven, J., Raven, J.C., & Court, J.H. (1998). Manual for Raven’s Progressive Matrices

and Vocabulary Scales (Raven Manual: Section 1: General Overview). Banbury

Road, Oxford: Oxford Psychologists Press Ltd.

Raven, J., Raven, J.C., & Court, J.H. (2000). Manual for Raven’s Progressive Matrices

and Vocabulary Scales (Raven Manual: Section 3: Standard Progressive

Matrices). San Antonio, Texas: Harcourt Assessment Inc.

Reber, A.S. (1995). Dictionary of Psychology. (2nd

Ed.). Ringwood, Victoria: Penguin

Group.

Riketta, M. (2008). The casual relation between job attitudes and performance: A meta-

analysis of panel studies. Journal of Applied Psychology, 93(2), 472-486.

Robbins, S.P., & Hunsaker, P.L. (2006). Training in interpersonal skills: Tips for

managing people at work (4th ed.). New Jersey: Pearson Prentice Hall.

Robbins, S.P., Waters-Marsh, T., Cacioppe, R. & Millet, B. (1994). Organisational

Behaviour: Concepts, Controversies and Applications, Australia: Prentice Hall.

Roethlisberger, F.J. & Dickson, W.J. (1939). Management and the Worker. Cambridge,

MA: Harvard University Press.

Rotenberry. P.F. & Moberg, P.J. (2007). Assessing the impact of job involvement on

performance. Management Research News, 30(3), 203-215.

Royal Australian Navy. (2011). Uniform Ranks. Available Online:

http://www.navy.gov.au/Uniform_Ranks.

Rusting, C.L. & Larson, R.J. (1997). Extraversion, neuroticism, and susceptibility to

positive and negative affect: A test between two theoretical models. Personality

and Individual Differences, 22(5), 607-612.

Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and

Personality, 9(3), 185–211.

Sarros, J.C., Gray, J., Densten, I.L., & Cooper, B. (2005). The organizational culture

profile revisited and revised: An Australian perspective, Australian Journal of

Management, 30(1), 159-182.

Schindlmayr, T. & Ong, P. (2001). Defence Personnel Environment Scan 2020.

Canberra: Directorate of Strategic Personnel Planning & Research.


Schmidt, F.L. & Hunter, J. (2004). General mental ability in the world of work:

Occupational attainment and job performance. Journal of Personality and Social

Psychology, 86(1), 162-173.

Scholl, R.W., Cooper, E.A., & McKenna, J.F. (1987). Referent selection in determining

equity perceptions: Differential effects on behavioral and attitudinal outcomes.

Personnel Psychology, 40(1), 113-124.

Shelton, H.W. (1920). Mutual rating: Contribution to the technique of participation.

Bulletin of the Taylor Society, 5(1), 59-67.

Singh, A.K. & Singh, A.P. (2009). Does personality predict organisational citizenship

behaviour among managerial personnel. Journal of the Indian Academy of


Smith, P.C., Kendall, L.M., & Hulin, C.L. (1969). The Measurement of Satisfaction in

Work and Retirement. Chicago, IL: Rand McNally.

Smith, C. A., Organ, D. W., & Near, J. P. (1983). Organizational citizenship behavior:

Its nature and antecedents. Journal of Applied Psychology, 68, 653-663.

Spearman, C. (1904). "General intelligence," objectively determined and measured.

American Journal of Psychology, 15(2), 201–293.

Tabachnick, B.G. & Fidell, L.S. (1996). Using Multivariate Statistics, (3rd

Ed.). New

York, NY: HarperCollins College Publishers.

Tabachnick, B.G. & Fidell, L.S. (2007). Using Multivariate Statistics, (5th

Ed.). Boston,

MA: Pearson Education, Inc.

Tett, R.P., Jackson, D.N. & Rothstein, M. (1991). Personality measures as predictors of

job performance: A meta-analytic review. Personnel Psychology, 44(4), 703-742.

Tett, R.P., Jackson, O.N., Rothstein, M., & Reddon, J.R. (1999). Meta-analysis of

bidirectional relations in personality-job performance research. Human

Performance, 12(1), 1-29.

Triandis, H.C. (1995). Individualism & Collectivism, Westview Press, Boulder CO.

Tupes, E.C. & Christal, R.E. (1992). Recurrent personality factors based on trait ratings.

Journal of Personality, 60(2), 225-251, (original 1961).

Ullman, J.B. (1996). Structural equation modeling. In B.G. Tabachnick & L.S. Fidell

Eds.), Using Multivariate Statistics, (3rd

Ed.), (Chapter 14, pp. 709-819). New

York, NY: HarperCollins College Publishers.

Van Dyne, L., Cummings, L.L., & McLean Parks, J.M. (1995). Extra-role behaviors: In

pursuit of construct and definitional clarity (A bridge over muddied waters). In

L.L. Cummings & B.M. Staw (Eds.), Research in Organizational Behavior (Vol

17, pp.215-285). Greenwich, CT: JAI Press.


Van Scotter, J.R. & Motowidlo, S.J. (1996). Interpersonal facilitation and job dedication

as separate facets of contextual performance. Journal of Applied Psychology,

81(5), 525-531.

Vigota-Gadot, E. (2007). Redrawing the boundaries of OCB? An empirical examination

of compulsory extra-role behavior in the workplace. Journal of Business and

Psychology, 21(3), 377-405.

Vroom, V.H. (1964). Work and motivation. New York, NY: Wiley.

Wallace, J.C., Johnson, P.D., Mathe, K., & Paul, J. (2011). Structural and psychological

empowerment climates, performance, and the moderating role of shared felt

accountability: A managerial perspective. Journal of Applied Psychology, 96(4),

840-850.

Warr, P. & Bourne, A. (1999). Factors influencing two types of congruence in

multirater judgements. Human Performance, 12(3/4), 183-210.

Warr, P. & Bourne, A. (2000). Associations between rating content and self-other

agreement in multi-source feedback. European Journal of Work and

Organizational Psychology, 9(3), 321-334.

Watson, D. & Clark, L.A. (1984). Negative affectivity: The disposition to experience

aversive emotional states. Psychological Bulletin, 96(3), 465-490.

Watson, D. & Clark, L.A. (1992a). Affects separable and inseparable: On the

hierarchical arrangement of the negative affects. Journal of Personality and Social

Psychology, 62(3), 489-505.

Watson, D. & Clark, L.A. (1992b). On traits and temperament: General and specific

factors of emotional experience and their relation to the five-factor model. Journal

of Personality, 60(2), 441-476.

Watson, D. & Clark, L.A. (1997). Measurement and mismeasurement of mood:

Recurrent and emergent issues. Journal of Personality Assessment, 68(2), 267-

296.

Watson, D., Clark, L.A., McIntyre, C.W., & Hamaker, S. (1992). Affect, personality,

and social activity. Journal of Personality and Social Psychology, 63(6), 1011-

1025.

Watson, D., Clark, L.A., & Tellegen, A. (1988). Development and validation of brief

measures of positive and negative affect: The PANAS scales. Journal of

Personality and Social Psychology, 54(6), 1063-1070.

Weiss, H.M. (2002). Deconstructing job satisfaction. Separating evaluations, beliefs and

affective experiences. Human Resource Management Review, 12(2), 173-194.

Weiss, H.M. & Cropanzano, R. (1996). Affective events theory: A theoretical

discussion of the structure, causes and consequences of affective experiences at

work. In B.M. Staw & L.L. Cummings (Eds.), Research in Organizational

Behaviour. Greenwich, CT: JAI Press.


Weiss, D.J., Dawis, R.V., England, G.W., & Lofquist, L.H. (1967). Manual for the

Minnesota Satisfaction Questionnaire. Minnesota: University of Minnesota.

Werner, J.M. (2000). Implications of OCB and contextual performance for human

resource management. Human Resource Management Review, 10(1), 3-24.

Williams, L.J. & Anderson, S.E. (1991). Job satisfaction and organizational

commitment as predictors of organizational citizenship and in-role behaviors.

Journal of Management, 17(3), 601-617.

Williams, S.B. & Leavitt, H.J. (1947). Group opinion as a predictor of military

leadership. Journal of Consulting Psychology, 11(6), 283-291.

Witt, L.A., Burke, L.A., Barrick, M.R., & Mount, M.K. (2002). The interactive effects

of conscientiousness and agreeableness on job performance. Journal of Applied

Psychology, 87(1), 164-169.


APPENDICES

Appendix A Organisational Citizenship Behaviour Scale Items from the 1983 Study

Conducted by Bateman and Organ (Organ et al., 2006, pp.245-246)

Appendix B Major Motivation Theories (Katzell & Thompson, 1990, pp.145-146)

Appendix C Description of the Revised NEO Personality Inventory (NEO PI-R)


Appendix D Royal Australian Navy Rank Structure

Appendix E Petty Officer Research Information and Informed Consent Forms

Appendix F Observer Research Information and Informed Consent Forms

Appendix G Behavioural Scale – Self-Report by Petty Officers

Appendix H Colour Coded Behavioural Scale – Self-Report by Petty Officers

Appendix I Behavioural Scale – Observer Report from Supervisors, Peers, and

Subordinates

Appendix J Positive and Negative Affect Scale (PANAS)

Appendix K Organisational Citizenship Behaviour Trial Questionnaire as Assessed

by Category Sponsors

Appendix L Category Sponsor Research Information Sheet

Appendix A: Organisational Citizenship Behaviour Scale Items from the 1983 Study

Conducted by Bateman and Organ (Organ et al., 2006, p245-246)

Scale Items

1 Comes up with new, original ideas for handling work.

2 Conscientiously follows organizational rules.

3 Trains or helps others to perform their jobs better.

4 Takes a personal interest in other employees.

5 Acts impulsively, on the spur of the moment.(R)

6 Has ups and downs in moods. (R)

7 Critically finds fault with other employees, (R)

8 Makes sure that things are neat, clean and orderly.

9 Tries to look busy doing nothing. (R)

10 Resists influence from others, including the boss. (R)

11 Acts cheerfully.

12 Expresses resentment at being given orders. (R)

13 Loses touch with things going on around him/her. (R)

14 Cooperates well with those around him/her.

15 Exhibits punctuality in arriving at work on time in the morning and after breaks.

16 Takes undeserved work breaks. (R)

17 Complains about insignificant things at work. (R)

18 Seeks others’ help when he/she needs it.

19 Makes positive statements about his/her immediate supervisor.

20 Makes constructive statements about the department.

21 Purposefully interferes with someone else doing their job. (R)

22 Exhibits dependability in carrying out his/her responsibilities.

23 Has people go to him/her for assistance.

24 Goes out of his/her way to protect others employees.

25 Goes out of his/her way to protect organizational property.

26 Exhibits annoyance with others. (R)

27 Exhibits poor quality work. (R)

28 Starts arguments with other employees. (R)

29 Talks about wanting to quit his/her job. (R)

30 Wastes material or harms organizational property. (R)

Appendix B: Major Motivation Theories (Katzell & Thompson, 1990, pp.145-146)

Endogenous Motivation Theories – Theories to Understand the Motivation Construct

Arousal/Activation

Theory

Arousal/activation theories focus on internal processes that

mediate the effects of conditions of work on performance.

Physiological and affective states are the two types of mediators

that have received the most attention.

Expectancy-Valence

Theory

People are motivated when they expect that effort will result in

good performance, which in turn will be instrumental in attaining

valued outcomes.

Equity Theory People are motivated by their need for fair treatment. Justice

consists of a balance between a worker’s inputs in a given situation

(e.g., ability, seniority) and its outcomes (e.g., money, promotions).

Equity exists when output/input ratios for the individual employee

and the reference source (e.g., co-worker, profession) are equal.

Attitude Theory People who have favourable attitudes toward their jobs, work,

and/or organizations will be more highly motivated to remain in

and perform their jobs. The principle of cognitive consistency also

implies that people will act in ways that accord with their attitudes.

Two major work-related attitudes are job satisfaction (affect

associated with one’s job) and job involvement (how important the

job is to the incumbent).

Intention/Goal Theory A person’s performance is determined by the goals to which he or

she is committed. The goals may be self-set or accepted from those

set by others. Intentions are cognitive representations of goals to

which the person is committed. People who are committed to

specific, hard goals perform at higher levels than people who have

easier or vaguer goals.

Attribution/Self-

efficacy Theory

Although attribution and self-efficacy represent two somewhat

different theoretical strands, they can be merged in their

implications for work motivation. Attribution theory is concerned

with explanations that people have for why particular events occur

or why people behave as they do. If people think that the causes of

their performance are stable, internal, and intentional, successful

performance will affect their self-efficacy beliefs favorably. People

with perceptions of greater self-efficacy and higher self-esteem are

more likely to have higher performance standards and goals, have

expectations of better performance, have more favorable job

attitudes, and show greater willingness to put forth effort on

challenging tasks.

B-2

Exogenous Motivation Theories – Theories that Increase/Decrease Motivation

Motive/Need

Theory

People have certain innate or acquired propensities to seek out or

avoid certain kinds of stimuli. These propensities, called motives or

needs, influence behavior and are major determinants of performance.

Various theories differ in content regarding the number of basic needs

or sets of needs are arranged in some hierarchical order.

Incentive/Reward

Theory

Incentive consist of features of the work situation (e.g., what the

supervisor says and does) that lead the workers to associate certain

forms of behavior (e.g., high quality of product) with a reward (e.g.,

praise). Disincentives are stimuli that conversely evoke avoidance, or

refraining, such as a company policy that docks pay when employees

are absent. Incentives are therefore important in attracting and holding

employees and in directing behavior. Rewards are stimuli that satisfy

one or more motives and therefore arouse positive psychological states

that serve to encourage and maintain the behavior that produced them.

Reinforcement

Theory

People are motivated to perform well when there have been positive

consequences of good performance. Conversely, ineffective behavior

should not be positively reinforced or should be punished. The effects of

reinforcement depend heavily on the schedule according to which

reinforcers are delivered Hence, more attention is devoted to schedules

than to the properties of the reinforcers..

Goal Theory The basic proposition of goal theory is that people will perform better if

goals are defined that are difficult, specific, and attractive. People need

feedback to continue to perform at high levels. Commitment to a goal

may be increased by money or another concrete reward or by

participating in setting the work goals.

Personal and

Material Resource

Theory

Constraints on workers’ abilities or opportunities to attain their work

goals are demotivating. In the extreme, such constraints can lead to

apathy or learned helplessness. Conversely, conditions that facilitate

goal attainment are positively motivating. These constraints and

facilitators can be personal (such as skill level) or material (such as

equipment).

Group and Norm

Theory

People are motivated to perform well when their work group facilitates

and approves of it. The dynamics of formal and informal work groups

often include the development of cohesiveness, the emergence of norms

regarding behavior, particularly about how much work is appropriate,

and the conformity of individual members to these norms. The work

group develops and maintains adherence to norms through the use of

social rewards and sanctions. Working in the presence of other group

members is itself a source of arousal, especially if the other members

are perceived as monitoring or evaluating one’s performance. People

are also prone to absorb the attitudes and behavioural dispositions of

other group members.

Sociotechnical

System Theory

People are motivated to perform well when the work system is designed

so that conditions for effective personal, social, and technological

functioning are harmonized. The work should be meaningful,

challenging, and diversified, and workers should have the skills,

autonomy, and resources to do it well.

(Source: Katzell & Thompson, 1990, pp.145-146)

Appendix C: Description of the Revised NEO Personality Inventory (NEO PI-R)


NEUROTICISM FACETS

N1: Anxiety

Anxious individuals are apprehensive, fearful, prone to worry, nervous, tense, and

jittery. The scale does not measure specific fears or phobias, but high scorers are more

likely to have such fears, as well as free-floating anxiety. Low scorers are calm and

relaxed; they do not dwell on things that might go wrong.

N2: Angry Hostility

Angry hostility represents the tendency to experience anger and related states such as

frustration and bitterness. This scale measures the individual's readiness to experience

anger; whether the anger is expressed depends on the individual's level of

Agreeableness. Note, however, that disagreeable people often score high on this scale.

Low scorers are easygoing and slow to anger.

N3: Depression

This scale measures normal individual differences and the tendency to experience

depressive affect. High scorers are prone to feelings of guilt, sadness, hopelessness, and

loneliness. They are easily discouraged and often dejected. Low scorers rarely

experience such emotions, but they are not necessarily cheerful and lighthearted -

characteristics that are associated instead with Extraversion.

N4: Self-Consciousness

The emotions of shame and embarrassment form the core of this facet of Neuroticism.

Self-conscious individuals are uncomfortable around others, sensitive to ridicule, and

prone to feelings of inferiority. Self-consciousness is akin to shyness and social

anxiety. Low scorers do not necessarily have poise or good social skills; they are

simply less disturbed by awkward social situations.

C-2

N5: lmpulsiveness

In the NEO-PI-R, impulsiveness refers to the inability to control cravings and urges.

Desires (e.g., for food, cigarettes, possessions) are perceived as being so strong that the

individual cannot resist them, although he or she may later regret the behavior. Low

scorers find it easier to resist such temptations, having a high tolerance for frustration. The

term impulsiveness is used by many theorists to refer to many different and unrelated

traits. NEO-PI-R impulsiveness should not be confused with spontaneity, risk-taking, or

rapid decision time.

N6: Vulnerability

The facet of Neuroticism is vulnerability to stress. Individuals who score high on this scale

feel unable to cope with stress, becoming dependent, hopeless, or panicked when facing

emergency situations. Low scorers perceive themselves as capable of handling themselves

in difficult situations.

EXTRAVERSION FACETS

El: Warmth

Warmth is the facet of Extraversion that is most relevant to issues of interpersonal

intimacy. Warm people are affectionate and friendly. They genuinely like people and

easily form close attachments to others. Low scorers are neither hostile nor necessarily

lacking in compassion, but they are more formal, reserved, and distant in manner than are

high scorers. Warmth is the facet of Extraversion out is closest to Agreeableness in

interpersonal space, but it is distinguished by a cordiality and heartiness that is not part of

Agreeableness.

E2: Gregariousness

A second aspect of Extraversion is gregariousness - the preference for other people’s

company. Gregarious people enjoy the company of others, and the more the merrier. Low

scorers on this scale tend to be loners who do not seek - or who even actively avoid - social

stimulation.

C-3

E3: Assertiveness

High scorers on this scale are dominant, forceful, and socially ascendant. They speak

without hesitation and often become group leaders. Low scorers prefer to keep in the

background and let others do the talking.

E4: Activity

A high activity score is seen in rapid tempo and vigorous movement, in a sense of energy,

and in a need to keep busy. Active people lead fast-paced lives. Low scorers are more

leisurely and relaxed in tempo, although they are not necessarily sluggish or lazy.

E5: Excitement Seeking

High scorers on this scale crave excitement and stimulation. They like bright colors and

noisy environments. Excitement seeking is akin to some aspects of sensation seeking. Low

scorers feel little need for thrills and prefer a life that high scorers might fine boring.

E6: Positive Emotions

The last facet of Extraversion assesses the tendency to experience positive emotions such

as joy, happiness, love, and excitement. High scorers on the positive emotions scale laugh

easily and often. They are cheerful and optimistic. Low scorers are not necessarily

unhappy; they are merely less exuberant and high-spirited. Research has shown that

happiness and life satisfaction are related to both Neuroticism and Extraversion and that

positive emotions is the facet of Extraversion most relevant to the prediction of happiness.

OPENNESS TO EXPERIENCE FACETS

01: Fantasy

Individuals who are open to fantasy have a vivid imagination and an active fantasy life.

They daydream not simply as an escape but as a way of creating for themselves an

interesting inner world. They elaborate and develop their fantasies and believe that

imagination contributes to a rich and creative life. Low scorers are more prosaic and prefer

to keep their minds on the task at hand.

C-4

02: Aesthetics

High scorers on this scale have a deep appreciation for art and beauty. They are moved by

poetry, absorbed in music, and intrigued by art. They need not have artistic talent, nor

even necessarily what most people would consider good taste, but for many of them, their

interest in the arts will lead them to develop a wider knowledge and appreciation than that

of the average individual. Low scorers are relatively insensitive to and uninterested in art

and beauty.

03: Feelings

Openness of feelings implies receptivity to one’s own inner feelings and emotions and the

evaluation of emotion as an important part of life. High scorers experience deeper and

more differentiated emotional states and feel both happiness and unhappiness more

intensely than do others. Low scorers have somewhat blunted affects and do not believe

that feelings states are of much importance.

04: Actions

Openness is seen behaviorally in the willingness to try different activities, go new places,

or eat, unusual foods. High scorers on this scale prefer novelty and variety to familiarity

and routine. Over time, they may engage in a series of different hobbies. Low scorers find

change difficult and prefer to stick with the tried-and-true.

05: Ideas

Intellectual curiosity is an aspect of Openness that has long been recognized. This trait is

seen not only in an active pursuit of intellectual interests for their own sake but also in

open-mindedness and a willingness to consider new, perhaps unconventional ideas. High

scorers enjoy both philosophical arguments and brain teasers. Openness to ideas does not

necessarily imply high intelligence, although it can contribute to the development of

intellectual potential. Low scorers on this scale have limited capacity and, if highly

intelligent, narrowly focus their resources on limited topics.

C-5

06: Values

Openness to values means the readiness to reexamine social, political, and religious values.

Closed individuals tend to accept authority and honor tradition and as a consequence are

generally conservative, regardless of political party affiliation. Openness to values can be

considered the opposite of dogmatism.

AGREEABLENESS FACETS

Al: Trust

High scorers on this scale have a disposition to believe that others are honest and well-

intentioned. Low scorers on this scale tend to be cynical and sceptical and to assume that

others may be dishonest or dangerous.

A2: Straightforwardness

Straightforward individuals are frank, sincere, and ingenious. Low scorers on this scale are

more willing to manipulate others through flattery, craftiness, or deception. They view

these tactics as necessary social skills and may regard more straightforward people as

naïve. When interpreting this scale (as well as other Agreeableness and Conscientiousness

scales), it is particularly important to recall that scores reflect standing relative to other

individuals. A low scorer on this scale is more likely to stretch the truth or to be guarded in

expressing his or her true feelings, but this should not be interpreted to mean that he or she

is a dishonest or manipulative person. In particular, this scale should not be regarded as a

lie scale, either for assessing the validity of the test itself or for making predictions about

honesty in employment or other settings.

A3: Altruism

High scorers on this scale have an active concern for others’ welfare as shown in

generosity, consideration of others, and a willingness to assist others in need of help. Low

scorers on this scale are somewhat more self-centered and are reluctant to get involved in

the problems of others.

C-6

A4: Compliance

This facet of Agreeableness concerns characteristic reactions to interpersonal conflict.

The high scorer tends to defer to others, to inhibit aggression, and to forgive and forget.

Compliant people are meek and mild. The low scorer is aggressive, prefers to compete

rather than cooperate, and has no reluctance to express anger when necessary.

A5: Modesty

High scorers on this scale are humble and self-effacing although they are not necessarily

lacking in self-confidence or self-esteem. Low scorers believe they are superior people

and may be considered conceited or arrogant by others. A pathological lack of modesty

is part of the clinical conception of narcissism.

A6: Tendermindedness

This facet scale measures attitudes of sympathy and concern for others. High scorers are

moved by others’ needs and emphasize the human side of social policies. Low scorers

are more hardheaded and less moved by appeals to pity. They would consider

themselves realists who make rational decisions based on cold logic.

CONSCIENTIOUSNESS FACETS

C1: Competence

Competence refers to the sense that one is capable, sensible, prudent, and effective.

High scorers on this scale feel well prepared to deal with life. Low scorers have a lower

opinion of their abilities and admit that they are often unprepared and inept. Of all the

Conscientiousness facets, competence is most highly associated with self-esteem and

internal locus of control.

C2: Order

High scorers on this scale are neat, tidy, and well-organised. They keep things in their

proper places. Low scorers are unable to get organised and describe themselves as

unmethodical. Carried to an extreme, high order might contribute to a compulsive

personality disorder.

C-7

C3: Dutifulness

In one sense, conscientiousness means “governed by conscience,” and that aspect of

Conscientiousness is assessed as dutifulness. High scorers on this scale adhere strictly to

their ethical principles and scrupulously fulfil their moral obligations. Low scorers are

more casual about such matters and may be somewhat undependable or unreliable.

C4: Achievement Striving

Individuals who score high on this facet have high aspiration levels and work hard to

achieve their goals. They are diligent and purposeful and have a sense of direction in

life. Very high scorers, however, may invest too much in their careers and become

workaholics. Low scorers are lackadaisical and perhaps even lazy. They are not driven

to succeed. They lack ambition and may seem aimless, but they are often perfectly

content with their low levels of achievement.

C5: Self-Discipline

Self-discipline refers to the ability to begin tasks and carry them through to completion

despite boredom and other distractions. High scorers have the ability to motivate

themselves to get the job done. Low scorers procrastinate in beginning chores and are

easily discouraged and eager to quit. Low self-discipline is easily confused with

impulsiveness - both are evident of poor self-control - but empirically they are distinct.

People high in impulsiveness cannot resist doing what they do not want themselves to

do - people low in self-discipline cannot force themselves to do what they want

themselves to do. The former requires an emotional stability; the latter, a degree of

motivation that they do not possess.

C6: Deliberation

The final facet of Conscientiousness is deliberation - the tendency to think carefully

before acting. High scorers on this facet are cautious and deliberate. Low scorers are

hasty and often speak or act without considering the consequences. At best, low scorers

are spontaneous and able to make snap decisions when necessary.

Appendix D: Royal Australian Navy Rank Structure

UNIFORM RANKS

Officers

Rank Abbreviation Badge

Admiral ADML

Vice Admiral VADM

Rear Admiral RADM

Commodore CDRE

Captain CAPT

Commander CMDR

Lieutenant Commander LCDR

Lieutenant LEUT

Sub Lieutenant SBLT

Acting Sub Lieutenant ASLT

Midshipman MIDN

D-2

Non-Commissioned Officers & Sailors

Rank Abbreviation Badge

Warrant Officer Navy WO-N

Warrant Officer WO

Chief Petty Officer CPO

Petty Officer PO

Leading Seaman LS

Able Seaman AB

Seaman SMN

HMAS Sirius and HMAS

Appendix E: Petty Officer Research Information and Informed Consent

Forms

Appendix F: Observer Research Information and Informed Consent Forms

Appendix G: Behavioural Scale – Self-Report by Petty Officers

G-6

Note:

Appendix H contains the colour code for the items which identify if they measure OCB,

interpersonal skills, job satisfaction, job satisfaction importance, job involvement.

Appendix H: Colour Coded Behavioural Scale– Self-Report by Petty Officers

H-6

Note: The colour coding system for the questionnaire is:

Pink OCB items

Red Interpersonal skills items

Green Job satisfaction items

Purple Job satisfaction importance items

Blue Job involvement items

Appendix I: Behavioural Scale – Observer Report by Supervisors, Peers, and

Subordinates

I-4

Item Summary

Items 1-24: Podsakoff et al (1990) OCB Items

Items 25-39: Interpersonal Skills Items

Items 40-60: Additional Category Sponsor OCB Items

Appendix J: Positive and Negative Affect Scale (PANAS)

J-2

Note:

Positive Affect was measured by items: 1, 3, 5, 9, 10, 12, 14, 16, 17, 19

Negative Affect was measured by items: 2, 4, 6, 7, 8, 11, 13, 15, 18, 20

Appendix K: Organisational Citizenship Behaviour Trial Questionnaire as

Assessed by Category Sponsors

Appendix L: Category Sponsor Research Information Sheet

Charles Sturt University Research Output · Predicting Organisational Citizenship Behaviour ii...

Documents

Transcript of Charles Sturt University Research Output · Predicting Organisational Citizenship Behaviour ii...